Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 111 appraisals
IEEE transactions on cybernetics
Knowledge Distillation and Reinforcement Learning in a Human–Machine Collaboration Delivery System With a Robotic Arm
Robotic arms are widely used in various aspects of human-robot collaboration. The primary goal of this study is to explore the usability of robotic arms for delivering objects to humans in dynamic environments. Traditional robotic arms often face limitations in path planning, such as difficulties adapting to dynamic environments and complex developmental processes. To overcome these challenges, this study employs reinforcement learning (RL) to train four models-the Approach RL Model, Delivery RL Model, Decision RL Model, and Merged Model-as alternatives to conventional path planning control. Typically, there exists a significant discrepancy between simulated data and real-world features. Although image segmentation can substantially reduce the gap between virtual and real environments, notable differences remain in hand features. Therefore, to further bridge the simulation-to-reality gap, this study applies CycleGAN to transform real hand features into virtual hand features, thereby enhancing the model's transferability. Experimental results show that the Decision RL Model achieved an accuracy of 99.17%, while the Merged Model achieved 99.92%. The proposed method effectively improves the stability and accuracy of human-robot collaboration in complex scenarios. Overall, this study validates the feasibility of integrating RL, image segmentation, and image translation techniques, offering a scalable and efficient task-solving solution for robotic arms in highly dynamic application domains.
3 Aug 2026
Read appraisal →The Knee
Artificial intelligence in the management of sports knee injuries: a narrative review.
INTRODUCTION: Sports-related knee injuries are common and debilitating, often leading to chronic pain, early osteoarthritis, and reduced performance. Artificial Intelligence (AI) has emerged as a promising tool to improve their prevention, diagnosis, prognosis, and rehabilitation. This review summarises current evidence on the clinical applications, limitations, and future directions of AI and machine learning in sports-related knee injuries. METHODS: A narrative review of PubMed, Embase, Medline and Web of Science was conducted, examining recent literature on AI-based models in musculoskeletal and sports medicine. The review was categorised into key domains: injury prediction and prevention, diagnostic imaging performance, AI-enabled clinical workflows, alongside postoperative and rehabilitation outcome modelling. RESULTS: AI algorithms demonstrate strong potential across the sports knee injury continuum. Predictive models analysing biomechanical and physiological data have achieved high area under the curve (AUC) values, in some cases above 0.90, in experimental and pilot setting when identifying athletes at risk of ACL rupture or overuse injuries, while machine learning approaches have been used to predict graft failure, revision surgery, and return-to-sport. However, most remain investigational rather than clinically deployable, with limited explainability, insufficient external validation, and training datasets that are often narrow or unrepresentative of broader athletic populations. CONCLUSION: AI has the potential to transform the management of sports-related knee injuries through more predictive, personalised, and precise care. However, wider clinical adoption will require multicentre validation, improved interpretability, and robust ethical and regulatory oversight. With further development, AI may enhance injury prevention, recovery, and improve long-term joint health outcomes in athletes.
3 Aug 2026
Read appraisal →Simulation in healthcare : journal of the Society for Simulation in Healthcare
Future of Biometric Technology in Healthcare Simulation
Biometric technology shows substantial promise to enhance patient safety and transform provider training. In patient care, biometric technology enables outpatient providers to optimize chronic disease by alerting providers to changes in patient conditions with real-time data. In provider training, biometric technology can inform training environments and curriculum. Certainly, there are hurdles to technology implementation including hardware fragility, cost, and software integration. Despite challenges, biometric technology has been effectively implemented in simulations and real patient settings. Looking forward, future developments focused on data security and privacy, standardizing data collection, and improving system compatibility will facilitate the wider adoption of biometric technology and improve health care quality, especially in resource-limited settings. Collaborative efforts among informatics experts, patient safety advocates, and health care providers are essential to overcome challenges and fully leverage biometric technology's potential in enhancing patient outcomes. The purpose of this commentary is to review current applications, identify key challenges, and offer recommendations to the simulation community for testing biometric technology, use in training programs, and facilitating systems integration to improve outcomes.
2 Aug 2026
Read appraisal →Bioelectrochemistry (Amsterdam, Netherlands)
From data to diagnosis: A comprehensive review of machine learning-driven wearable sensors in healthcare.
The integration of machine learning (ML) with advanced wearable sensor technologies is revolutionizing healthcare by enabling real-time, intelligent monitoring of physiological parameters such as electrocardiogram (ECG), blood glucose, and respiratory patterns. This review systematically examines the transformative potential of ML-driven biosensors across three core domains: health monitoring, early disease detection, and precision medicine. Key technological advancements-including self-optimizing sensor networks, explainable AI (XAI) architectures, and edge-computing-enabled miniaturized devices-are critically evaluated. Despite rapid progress, the translation of these technologies into clinical practice faces significant challenges, such as data standardization, algorithmic interpretability, privacy concerns, and regulatory hurdles. This paper also discusses emerging trends, including federated learning, quantum machine learning, and neural interfaces, which hold promise for overcoming these barriers. By addressing these challenges and leveraging ongoing interdisciplinary collaborations, ML-enhanced wearable systems are poised to redefine personalized medicine and proactive healthcare delivery on a global scale.
2 Aug 2026
Read appraisal →Ultrasound in medicine & biology
A Systematic Review: The Application of Attention Mechanisms in Medical Ultrasound Image Processing
Among numerous medical imaging modalities, ultrasound imaging is one of the most commonly used diagnostic methods in clinical practice. However, ultrasound diagnosis heavily relies on physician experience, and diagnostic results often lack reproducibility. In recent years, with the rapid development of artificial intelligence technology, which provides new impetus for the automated medical ultrasound image processing and analysis. Among the numerous deep learning approaches proposed, attention mechanisms have become a key component for improving network robustness to cope with challenges in ultrasound imaging, such as low contrast, blurred boundaries, and variable object morphologies. This paper systematically reviews the attention mechanisms employed in medical ultrasound image analysis, which can be roughly divided these mechanisms into three categories based on the differences in feature focus dimensions: channel attention, spatial attention, and hybrid attention. Most importantly, we not only summarized the application scenarios and effectiveness of various attention mechanisms but also analyzed the potential challenges faced in the future.
2 Aug 2026
Read appraisal →Medical physics
Reference-Free large language model agents for physician-guided radiotherapy treatment planning
BACKGROUND: Large language models (LLMs) have recently demonstrated exceptional capabilities, offering the potential to streamline workflows and enhance efficiency across diverse tasks. However, their application in domain-specific areas, such as radiotherapy treatment planning, remains challenging due to the lack of publicly available, specialized knowledge required for these tasks. PURPOSE: To investigate the critical architectural and functional components necessary to adapt an off-the-shelf large language model into a clinically viable planning agent for inverse treatment planning in intensity-modulated radiation therapy (IMRT). MATERIALS/METHODS: Twenty head-and-neck (HN) cancer patients who received IMRT at our institution were retrospectively collected under IRB approval. The LLM agent was implemented to directly interact with the clinical treatment planning system (TPS) to iteratively extract intermediate plan states and propose new constraint values to guide inverse optimization. Its decision-making was informed by real-time plan evaluations and prior optimization outcomes, enabling adaptive refinement of planning strategies across iterations. The agent was equipped with three core capabilities: comprehension of clinical planning objectives, contextual understanding of the optimization environment, and arithmetic proficiency for quantitative reasoning. Treatment planning was conducted in a reference-free inference setting, wherein the LLM operated without prior exposure to manually generated treatment plans and without any fine-tuning or task-specific training. LLM-generated treatment plans were compared with clinically approved plans created by certified dosimetrists. Key dosimetric endpoints were evaluated and statistically analyzed. Paired comparisons between LLM-generated and clinical plans were performed using the Wilcoxon signed-rank test. RESULTS: LLM-generated plans achieved comparable organ-at-risk (OAR) sparing relative to clinical plans, while demonstrating improved hot spot control (Dmax: 106.5% vs. 108.8%, p < 0.05) and superior conformity (conformity index: 1.18 vs. 1.39, p < 0.05 for boost PTV; 1.82 vs. 1.88, p = 0.47 for primary PTV). CONCLUSIONS: This study demonstrates the feasibility of a reference-free, LLM-driven workflow for automated IMRT treatment planning in a commercial TPS. The proposed approach provides a generalizable solution that could reduce planning variability and support broader adoption of AI-based planning strategies.
2 Aug 2026
Read appraisal →Rheumatic diseases clinics of North America
Toward Artificial Intelligence-driven Clinical Decision Support Tools in Rheumatology
Clinical decision support systems (CDSS) have the potential to enhance rheumatology practice by assisting with differential diagnosis, treatment decisions, and predicting patient outcomes. Rheumatic conditions are complex diseases largely diagnosed clinically rather than with a predefined set of clinical, laboratory, or imaging findings. Newer artificial intelligence (AI)-driven CDSS, known as machine-learning-based systems, may be better suited for applications to rheumatic conditions compared to knowledge-based systems, which rely on a set of predefined rules. This report reviews existing studies on CDSS in rheumatology and highlights the potential benefits and pitfalls of incorporating AI-driven CDSS into rheumatology practice.
1 Aug 2026
Read appraisal →Surgery
Artificial intelligence to revolutionize surgical decision-making is still just around the corner.
Medical artificial intelligence, especially large language models, has engendered both excitement and unease across the medical community, promising improved surgical diagnostic accuracy, perioperative decision-making, and patient safety amidst concerns of considerable bias and its implications on the future of human expertise. The evolution of artificial intelligence clinical decision support originated in the surgical field through early systems like AAPHelp, which leveraged Bayesian reasoning to diagnose acute abdominal pain. Despite the initial promise of artificial intelligence clinical decision support, attempts to develop more robust diagnostic tools largely fell short in the 1980s, with instruments unable to adequately diagnose and manage complex presentations, a lack of transparency in their decision-making processes, and limitations in scope. It was not until the 2010s that clinical decision support began to show renewed potential as a meaningful adjunct in surgical practice, largely driven by advancements in machine learning techniques and wider access to large clinical data sets. New artificial intelligence models, particularly those utilizing large language models, now demonstrate impressive capabilities, from predicting the risk of postoperative complications for individual patients to streamlining clinical documentation and beyond. However, challenges persist regarding transparency, bias, reliability, and integration into clinical workflows. Despite these hurdles, artificial intelligence-based tools like large language models represent an exciting new chapter in surgical decision-making. Built on a long history of clinical decision support systems in surgery, these technologies hold great promise to meaningfully augment surgical practice and improve care for patients.
1 Aug 2026
Read appraisal →Indian journal of ophthalmology
Internet of things in assistive technology for people with visual impairment: A scoping review
Assistive technology (AT) can significantly enhance functional abilities among persons with disabilities. Recently, a rapid advance in computer and Internet of Things (IoT)-based AT has been observed with the integration of Artificial Intelligence. The present scoping review explores a range of IoT-based AT for vision impairments, focusing on their functioning, reliability, and identifying the existing gaps and opportunities for future enhancement. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. The search was conducted for the articles published between 2020 and March 2025. After the initial screening of both titles and abstracts of 705 articles, 18 full-text articles were selected for the review. It was found that the majority of the papers (38%) are related to smart canes/sticks. The findings indicate that several studies use Raspberry Pi and Arduino platform to make products cost-effective and flexible options for real-time processing. The accuracy of these devices ranges from 87.8% in navigation to 93% in voice control and 100% in object identification. Further, challenges such as usability testing, hardware limitations, dataset constraints, and environmental adaptability remain a matter of concern which need to be addressed in future innovation. Incorporating IoT technologies into vision-assistive devices has substantially improved accessibility. Future research should focus on enhancing real-time performance, AI-driven decision-making, and user-centric designs.
1 Aug 2026
Read appraisal →Journal of pharmacokinetics and pharmacodynamics
Diffusion models for virtual populations and pharmacometric simulations
To evaluate diffusion model-based artificial intelligence approaches for generating virtual populations with physiological determinants of drug dosing (PDODD) and pharmacokinetic (PK) profiles. A denoising diffusion probabilistic model (DDPM) was applied to a 31-variable dataset of PDODD covariates (18 continuous, 13 binary) from the National Health and Nutrition Examination Survey and compared to a tabular variational autoencoder (TVAE). For nivolumab PK data (12,000 patients, 13 time points, 5 covariates), sequence-based diffusion model (SDM) and a time-aware diffusion model (TDM) with temporal self-attention were evaluated. The predictive performance of the TDM was evaluated by imputing masked time points. All models were trained and tested on 80%:20% partitions of the data using univariate, bivariate, and multivariate distributional similarity metrics. The diffusion model satisfactorily approximated the univariate distributions of continuous PDODD biomarkers (mean Kolmogorov-Smirnov D-statistic, KSD = 0.014), disease status frequencies (mean absolute error, MAE = 0.31%), and preserved bivariate correlations (MAE = 0.033). DDPM outperformed TVAE for categorical variables (0.31% vs. 1.07% MAE) and correlation (0.033 vs. 0.091 MAE). For nivolumab PK, SDM has KSD of 0.047 and a relative error of 1.36%. TDM accurately imputed missing PK timepoints (KSD = 0.014), reconstructing masked Day 1, Peak concentration (Cmax) Dose-9, and Terminal phase concentrations with MAE of 0.43%, 0.25%, and 0.76%, and correlations ≥ 0.999. Diffusion models demonstrated strong performance in generating cross-sectional PK covariate data and longitudinal PK profiles, capturing complex distributional and temporal dependencies. Diffusion-based approaches provide a flexible and robust framework for virtual simulations in pharmacometrics.
1 Aug 2026
Read appraisal →Scientific reports
Safety design guidelines for clinician-AI interaction in computer-aided diagnosis systems using system-theoretic framework with explainability validation
Ensuring the safety of Artificial Intelligence-enabled Computer-Aided Diagnosis systems is critical because diagnostic errors can have serious consequences for patient care. However, existing regulatory and risk management frameworks often do not sufficiently address the complex socio-technical interactions between clinicians and Artificial Intelligent systems, leaving key human-centered safety challenges underexplored. This paper presents a systematic and human-centered approach to deriving safety design guidelines for clinician-Artificial Intelligence interaction in Computer-Aided Diagnosis systems using System-Theoretic Process Analysis. Through this analysis, we identify critical hazards associated with clinician-Artificial Intelligence collaboration, including automation bias on system recommendations, and misinterpretation of explanations. Based on the identified unsafe control actions, we formulate a set of actionable and traceable safety design guidelines that promote transparency, coherent explanations, and calibrated trust in Artificial Intelligence-assisted decision-making. To bridge safety analysis and system design, the proposed guidelines are operationalized within a Computer-Aided Diagnosis interaction framework. The framework includes a safety-oriented Graphical User Interface that integrates multiple explanation methods and interactive mechanisms to promote clinician engagement. Furthermore, we introduce a safety-oriented evaluation approach that uses consistency across multiple explanation methods as a quantitative indicator of potentially unreliable or ambiguous explanations. By linking System-Theoretic Process Analysis, interaction design, and explainability evaluation, this work provides a unified and reusable framework for improving the safety and reliability of Artificial Intelligence-driven Computer-Aided Diagnosis systems.
31 July 2026
Read appraisal →Proceedings of the National Academy of Sciences of the United States of America
The backfiring effect of weak AI safety regulation
Recent policy proposals aim to improve the safety of general-purpose AI, but there is little understanding of the efficacy of different regulatory approaches. We present a strategic model that explores interactions between safety regulation, general-purpose AI technology creators, and domain specialists-those who adapt the technology for specific applications. Our analysis examines how regulatory measures targeting different parts of the AI development chain affect the outcome of this game. Our model assumes AI technology is characterized by two key attributes: safety and performance. The regulator first sets a minimum safety requirement that applies to one or both players. The general-purpose creator then invests in the technology, establishing its initial safety and performance levels. Next, domain specialists refine the AI for their use cases, updating the safety and performance levels and taking the product to market. Resulting revenue is shared between the specialist and generalist. Our analysis reveals two insights: first, weak safety regulation imposed predominantly on domain specialists can backfire. While it might seem logical to regulate AI use cases, our analysis shows that weak regulations targeting domain specialists alone can reduce safety in a large class of parameterizations. Second, in contrast to the previous finding, we observe that stronger, well-placed regulation can mutually benefit all players. When regulators impose appropriate safety standards on both general-purpose AI creators and domain specialists, the regulation can function as a commitment device, leading to safety and performance gains, surpassing what is achievable under no regulation or regulating only one player.
28 July 2026
Read appraisal →Journal of robotic surgery
From robot assistance to surgical intelligence: global research trends and emerging frontiers of artificial intelligence-enhanced robotic surgery in urology.
To systematically characterize global publication trends, collaboration patterns, intellectual foundations, research hotspots, and emerging frontiers in artificial intelligence (AI)-enhanced robotic and robot-assisted surgery in urology. Publications were retrieved from the Web of Science Core Collection (WoSCC) using a topic search strategy that included three keyword groups: robot-assisted surgery, urologic diseases or procedures, and AI-related technologies. English-language articles and reviews were included. Meeting abstracts, conference proceedings, editorials, letters, non-English publications, studies unrelated to urologic robotic surgery, and studies without substantive AI or intelligent algorithmic content were excluded. RStudio was used for descriptive statistics and annual publication trend visualization. VOSviewer was used for auxiliary bibliometric network construction and visualization. CiteSpace was used to construct collaboration networks, co-citation networks, keyword co-occurrence maps, cluster maps, timeline and time-zone maps, and citation burst maps. A total of 401 records were initially retrieved, and 253 publications were finally included after screening by document type, language, and topical relevance. These comprised 199 articles (78.66%) and 54 reviews (21.34%). Publications in this field began in 2004 and increased rapidly after 2018, reaching a peak of 53 publications in 2025. Because 2026 was an incomplete retrieval year, only 23 publications were recorded. Italy, the United States, the Netherlands, China, and Japan were the leading contributing countries. The University of Turin, Azienda Ospedaliero-Universitaria San Luigi Gonzaga, IRCCS Fondazione del Piemonte per l'Oncologia, the Netherlands Cancer Institute, and Leiden University Medical Center were the most productive institutions. Keyword clustering showed that the major research hotspots included renal cell carcinoma, augmented reality, image-guided surgery, indocyanine green, machine learning, registration, deep learning, and bladder cancer. AI-enhanced robotic surgery in urology has evolved from early research on registration, navigation, and image-guided surgery toward a surgical intelligence stage characterized by augmented reality, three-dimensional reconstruction, machine learning-based prediction, deep learning-based segmentation, surgical video understanding, and skill assessment. Future studies should prioritize multicenter standardized datasets, sharing of intraoperative video and robotic platform data, model interpretability, prospective validation, and integration into real-world clinical workflows.
27 July 2026
Read appraisal →Scientific data
Anonymized but Useful Synthetic Tabular Health Data for AI based Fall Risk Assessment
Artificial Intelligence (AI) bears potential for improving health care, but this depends on the availability of open-access, realistic, and useful data. To facilitate AI model development in health care we release SynTabFall, a novel synthetic dataset for fall risk assessment. With a total of 745,380 samples and 44 attributes such as demographics, diseases, mobility and cognition related risk factors, this tabular dataset allows for training fall risk prediction models without access to the original patient data. Models trained on our synthetic dataset can reach predictive performance scores in fall risk assessment which are on par with models trained on real data. To support others in sharing data we also describe a process that was developed over multiple years in one of Germany's largest hospitals in close collaboration between data protection officers, health care staff, informaticians and AI engineers. The proposed data sharing approach combines established methods for anonymization and modern generative AI (genAI) methods for synthesizing tabular data and allows for sharing health care data responsibly without sacrificing its utility. We release the synthetic fall risk dataset along with the software developed for synthetic data generation and evaluation.
26 July 2026
Read appraisal →PloS one
Textual overlap rather than domain alignment: A comparative study of fine-tuning strategies for specialised machine translation with large language models
General-purpose large language models (LLMs) may struggle in specialised machine translation, but the conditions under which fine-tuning improves translation performance remain unclear. This study compares full-parameter fine-tuning (FPFT) and parameter-efficient fine-tuning (PEFT) for Chinese-English political discourse translation using a purpose-built corpus and the Qwen3-14B model. Translation performance was assessed on three 50-item test sets using BLEU, ROUGE-L F1, METEOR, and BERTScore F1, together with BLEU pass-rate likelihood-ratio G2 tests, paired t-tests, and paired Cohen's dz for item-level score differences. The results reveal a clear contrast between unseen in-domain evaluation, maximum-overlap benchmarking, and semantically related but non-fine-tuned evaluation. On Test Set A and Test Set C, neither fine-tuning strategy produced a statistically significant BLEU pass-rate advantage over the base model, and paired tests across the continuous metrics did not show consistent fine-tuning gains. On Test Set B, which was sampled from the fine-tuning corpus, both fine-tuned models substantially outperformed the base model across all four metrics, with FPFT achieving the highest scores and PEFT providing a more computationally efficient alternative. These findings indicate that textual overlap between training and deployment data, rather than broad domain similarity alone, strongly conditions the observed benefit of fine-tuning. The study offers an empirically grounded framework for selecting fine-tuning strategies in specialised machine translation.
22 July 2026
Read appraisal →JMIR human factors
Video-Algorithmic Patient Monitoring in Mental Health Inpatient Settings: Qualitative Study of Patient or Consumer, Clinician, and Vendor Perspectives
BACKGROUND: Video-algorithmic patient monitoring (VAPM) combines remote, noncontact sensors and algorithmic analysis and is increasingly trialed in acute psychiatric and other care settings. While promoted for improving safety and reducing risk, it raises ethical concerns regarding safety, privacy and surveillance. Little is known about how those encountering VAPM in mental health care contexts anticipate its use and potential impacts, including where it has not yet been implemented. OBJECTIVE: This study aimed to explore the views of patients or mental health consumers, specialized mental health nurses and nurse academics, hospital managers, and technology vendors regarding the appropriateness and anticipated implications of VAPM in mental health inpatient care. METHODS: This qualitative study identified key stakeholders in Australia via networking techniques for participation in a deliberative workshop. A deliberative workshop was held, and the workshop discussion was audio-recorded, transcribed, and thematically analyzed, consistent with methods in health technology research, which enable exploration of different viewpoints, including convergences and divergences across stakeholder groups. RESULTS: In total, 16 stakeholders participated, exploring themes concerning (1) contestation over the rationale for VAPM in mental health settings, (2) VAPM reshaping care and relationships, (3) perceived harms of VAPM, (4) perceived observational support for safety and reduced disruption, (5) serious privacy implications of VAPM, (6) the need for appropriate governance, and (7) the potential for VAPM to transform, not augment, service delivery. General views differed across groups. Patients or service users expressed concerns about privacy, coercion, and the potential to intensify stigma. Mental health nurses were cautious but interested in possible benefits for safety and suicide prevention. Hospital managers and technology vendors largely emphasized safety gains. CONCLUSIONS: The findings suggest that the anticipated risks of VAPM are primarily experienced subjectively, as infringements on privacy, dignity, and trust, while purported benefits remain largely untested and unquantified. From a utilitarian perspective, direct comparison is therefore difficult-the risks are set out in the anticipated experiences of those with lived experience, and the benefits remain hypothetical. From this view, robust, independent evidence of real-world outcomes is required. Yet, for some participants, the very premise of such calculation was rejected, with privacy, dignity, and trust regarded as nonnegotiable, rather than items for trade-off. If VAPM is to be pursued at all, it should proceed only with extreme caution, with transparent evidence of outcomes, and with meaningful participation from those whose lives and care are most directly impacted.
22 July 2026
Read appraisal →Current biology : CB
Looking to the brain to improve energy efficiency of AI
Modern artificial intelligence (AI) systems have achieved remarkable capabilities, but at an extraordinary energy cost. Training and running large-scale models can consume vast resources, posing environmental, economic, and societal challenges. In contrast, biological brains perform lifelong learning, adaptive control, and flexible reasoning using orders of magnitude less energy for learning and adaptation over a lifetime. What accounts for this difference - and how can it guide future AI development? In this review, we identify key biological principles that support energy-efficient capacities in biological brains, and consider how they might inform the design of more sustainable artificial systems. We organize our analysis around three domains: architectural constraints, signaling strategies, and learning algorithms. In each domain, we discuss concrete observations from biology, from cell to circuit to cognitive level, and describe how current and emerging AI systems mirror or diverge from these motifs. One striking feature of biological energy optimization is often overlooked: that brains are remarkably stable in their energy usage across heterogeneous modes, suggesting they may minimize energy needs during active environmental processing through maximizing the utility of 'rest-like' background processes. Overall, rather than advocating for biomimicry for its own sake, we argue for biologically informed engineering. Understanding how natural systems minimize energetic cost while maximizing flexibility may help us build AI that is not only powerful, but also efficient, equitable, and environmentally responsible.
21 July 2026
Read appraisal →Journal of robotic surgery
Robotic versus laparoscopic and open surgery for endometrial cancer: a systematic review of randomized trials and pooled analysis of conversion rates
To summarize randomized evidence assessing robotic surgery in relation to conventional laparoscopic and open abdominal approaches for endometrial cancer treatment, with particular attention to perioperative outcomes and conversion to open surgery. Randomized controlled trials evaluating robot-assisted surgical treatment of endometrial cancer were identified across major biomedical databases up to December 2025. Eligible studies compared the robotic approach against laparoscopic or open abdominal surgery. Quantitative pooling was undertaken only when outcome reporting was sufficiently consistent across studies. Eight randomized trials including 647 patients met the inclusion criteria. Overall, 322 patients underwent robotic surgery, 244 conventional laparoscopy, and 81 laparotomy. Most perioperative endpoints were reported heterogeneously, limiting formal pooling. Operative time varied across trials when robotics was compared with laparoscopy and was generally longer than laparotomy. Intraoperative blood loss and postoperative hospitalization did not show consistent differences between the two minimally invasive approaches. Compared with laparotomy, the robotic approach was linked to reduced postoperative stay. Conversion to laparotomy occurred less frequently after robotic surgery than after laparoscopy (0.7% vs. 8.4%; OR 0.17; p = .03). Complication reporting was inconsistent, although trials comparing robotics with laparotomy generally favored the robotic approach. Direct procedural costs were higher for robotics, whereas indirect costs favored robotics in the single study evaluating them. The robotic approach resulted in a lower need for open conversion compared with conventional laparoscopy. Other perioperative outcomes appeared broadly comparable between the two minimally invasive approaches, while comparisons with laparotomy suggested shorter hospital stay and fewer postoperative complications, although these findings should be interpreted cautiously because of the limited and heterogeneous randomized evidence.
21 July 2026
Read appraisal →Journal of robotic surgery
Safety and feasibility of robotic telesurgery in urological procedures: a systematic review and meta-analysis of clinical data
Robotic telesurgery has emerged as a potential solution to expand access to specialized surgical care; however, clinical evidence remains fragmented across procedures and robotic platforms. We conducted this meta-analysis to assess the feasibility and safety of robotic telesurgery across urological procedures. An electronic search was conducted on PubMed, Scopus, Web of Science (WoS) and the Cochrane Library from inception through April 2026 according to PRISMA guidelines. Clinical studies reporting robotic telesurgery in urology were included regardless of study design. The primary outcome of interest was procedural success, defined as completion without conversion to a local surgeon or an alternative surgical approach. Secondary outcomes included complication rates and round-trip latency (RTL). Nineteen studies involving 164 patients were included. The pooled procedural success rate was 0.92 (95% CI: 0.86-0.96). No conversions to open or laparoscopic surgery were reported across the included studies. The pooled complication rate was 0.12 (95% CI: 0.07-0.20), with events occurring in a limited number of studies. RTL data were available from 14 studies, with a pooled mean latency of 123.30 milliseconds (ms) (95% CI: 81.76-164.83) demonstrating substantial heterogeneity (I² = 99.99%). Robotic telesurgery in urology demonstrates excellent technical feasibility and a favorable short-term safety profile in carefully selected cases. However, variability in network performance and procedure-specific requirements remain important considerations. Further prospective studies with standardized reporting and long-term outcome assessment are needed before widespread clinical adoption can be recommended.
21 July 2026
Read appraisal →Journal of robotic surgery
Mapping the evolution of deep learning and computer vision in robotic surgery: a bibliometric analysis of surgical video intelligence, instrument perception, and clinical translation.
Deep learning and computer vision are increasingly embedded in robotic surgery, yet the development and translational direction of this research domain remain incompletely characterized. We conducted a bibliometric and visualization analysis of publications retrieved from the Web of Science Core Collection using Bibliometrix/Biblioshiny, VOSviewer, and CiteSpace. A total of 1,186 documents published between 2010 and 2026 across 356 sources were included. Scientific output increased rapidly, with an annual growth rate of 16.09% and a peak of 216 publications in 2025. The field involved 5,296 authors, and international collaboration accounted for 30.69% of publications. IEEE Robotics and Automation Letters was the most productive and locally influential source. China and the United States were the leading contributors, with China showing the most rapid recent expansion and the United States retaining the highest citation impact. Citation-burst and keyword analyses identified U-Net, residual learning, transformer architectures, and foundation-model-enabled segmentation as major methodological drivers. The conceptual structure evolved from image guidance, registration, and navigation toward surgical video intelligence, instrument perception, workflow understanding, autonomous assistance, and clinical translation. Instrument perception emerged as a central link between algorithmic development and operative application. Future progress will require diverse multi-institutional datasets, external and prospective validation, integrated scene understanding, and rigorous evaluation of intelligent assistance within real robotic surgical workflows.
21 July 2026
Read appraisal →PloS one
A novel dynamic Optuna hybrid Harris Hawks Optimization approach for classification of CAD
Coronary Artery Disease (CAD) is a leading cause of mortality worldwide and is primarily associated with atherosclerotic plaque formation, resulting in coronary artery stenosis. The accurate prediction of CAD using deep learning models is often constrained by the limitations of conventional optimization techniques, including premature convergence and limited adaptability of the models. To address these challenges, this study proposes a Dynamic Optuna Hybrid Harris Hawks Optimization (Dynamic Optuna H-HHO) framework to enhance the performance of deep learning-based CAD prediction models. The proposed approach integrates dynamic parameter adjustment, adaptive escape energy mechanisms, and Optuna-based hyperparameter tuning. The framework was applied to optimize several deep-learning classifiers, including ResNet-50, VGG-16, InceptionV3, and MobileNet, using a coronary artery stenosis dataset. The performance was evaluated through a comparative analysis with models optimized using the conventional Hybrid Harris Hawks Optimization (H-HHO) algorithm. The experimental results indicate that the proposed Dynamic Optuna H-HHO framework consistently improves the predictive accuracy across all evaluated models. InceptionV3 achieved the highest accuracy of 97.9%, followed by MobileNet with 97.6%, compared with the maximum accuracy of 82.46% obtained using traditional HHO-based optimization. By combining adaptive optimization strategies with automated hyperparameter tuning, the proposed framework provides a robust and scalable solution for improving the accuracy of coronary artery disease prediction.
20 July 2026
Read appraisal →Medicine
Exploring the role of generative artificial intelligence in enhancing clinical skills training: A bibliometric analysis
BACKGROUND: Traditional clinical skills training faces challenges such as limited standardized patient resources, high costs of simulation equipment, and restricted opportunities for repeated practice. Generative artificial intelligence (GAI), particularly large language models, has attracted increasing attention as a potential tool for simulation, feedback, and adaptive learning in medical education. However, the research landscape and thematic development of GAI in clinical skills training remain insufficiently mapped. METHODS: A bibliometric analysis was conducted using literature retrieved from the Web of Science Core Collection from January 1, 2011 to April 1, 2025. VOSviewer, the Bibliometrix R package, and CiteSpace were used to analyze publication trends, country and institutional contributions, journal distribution, collaboration networks, keyword co-occurrence, co-cited references, and citation bursts. RESULTS: A total of 322 publications were included. Research activity remained limited before 2023 but increased rapidly thereafter. The United States contributed the largest number of publications, followed by China and India. Major contributing institutions included the National University of Singapore, Gazi University, and Nova Southeastern University. Frequently publishing journals included JMIR Medical Education, Medical Teacher, and BMC Medical Education. Keyword and thematic analyses showed that current research attention is mainly concentrated on ChatGPT, large language models, natural language processing, clinical reasoning simulation, personalized learning, virtual patient interaction, and ethical governance. CONCLUSIONS: Research on GAI in clinical skills training is in an early but rapidly expanding stage, with growing scholarly attention to language-model-driven educational applications and related ethical issues. The bibliometric findings reflect research activity, knowledge structure, and thematic priorities rather than direct evidence of educational effectiveness. Future studies should adopt rigorous empirical designs and standardized evaluation frameworks to assess the effectiveness, safety, and appropriate boundaries of GAI applications in clinical skills training.
19 July 2026
Read appraisal →JMIR cancer
Internet of Things-Enhanced Mathematical Oncology: Conceptual Framework for Adaptive Cancer Care Modeling
The Internet of Things (IoT) is transforming various industries, including health care. IoT-based systems are increasingly prevalent in consumer health applications, while intelligent or smart devices equipped with sophisticated sensors are gaining recognition for their potential to improve clinical care practice and decision-making. Cancer care is a particularly promising area for IoT applications, enabling real-time and personalized interventions. However, empirical research on the effects of IoT in this field is limited due to the complexities inherent in cancer as a dynamic disease and the paucity of IoT-generated data available for research. This presents an opportunity to apply mathematical modeling to understand the effects of IoT under various scenarios. These analytical and "in silico" mathematical approaches are instrumental with limited data. Such models support the analysis of treatment uncertainty and patient response while balancing patient preferences, clinical outcomes, and system-level constraints. Grounded in mathematical oncology and health informatics, this paper proposes a conceptual framework that integrates real-time IoT data as dynamic inputs into adaptive mathematical models to simulate cancer dynamics. By exploring applications across multiple levels of analysis, the study demonstrates how IoT-enhanced mathematical models could inform implementation and optimize oncology services, addressing a critical gap in current research.
15 July 2026
Read appraisal →Health care management science
An online algorithm for integrated scheduling of pre-treatment and treatment appointments in radiotherapy using deep reinforcement learning
Scheduling patients for radiotherapy within the oncology department is vital for ensuring timely and effective treatment. The patient's therapy consists of a sequence of pre-treatment consultations, i.e. preliminary investigations to define the required treatment, followed by the radiotherapy treatment sessions on linear accelerators. Timely scheduling of pre-treatment consultations is crucial to prevent unnecessary delays in radiotherapy treatment, thereby impacting patient health. Therefore, we suggest integrating both scheduling problems. We employ a Reinforcement Learning-based scheduling technique to solve the multi-appointment scheduling problem in an online fashion. We design a reward function comprising three components to guide the agent towards learning an effective scheduling policy involving multiple appointments. Additionally, we compare our proposed model with benchmark online heuristics. We further provide insight into the reward function design, which steers the agent to learn a feasible policy that leads to high-quality solutions. Experiments show that our algorithm outperforms the benchmark heuristics.
12 July 2026
Read appraisal →Science advances
A multifunctional porous interface bridging 3D architected electronics with skin
Skin-integrated flexible electronics are rapidly advancing from short-term use to continuous, long-term wear to meet the demands of health monitoring, wound healing, and skin disease treatment. While three-dimensionally (3D) architected devices offer sensing capabilities and outstanding mechanical performances, the development of a conformal interface between these 3D electronic devices and skin for long-term comfortable wear remains challenging due to their geometric complexity and mechanical fragility. Here, we introduce a multifunctional porous interface that bridges 3D flexible electronics with skin through engineered microporous networks. Such a porous design enhances elastic stretchability and electrical reliability while providing moisture permeability, thermal buffering, impact mitigation, and drug delivery capability. Stiffness reduction and localized pore-wall-shell buckling synergistically underpin the outstanding mechanical resilience under large deformations. The proposed multifunctional porous interface allows development of a 3D closed-loop wound administration patch capable of multimodal sensing and on-demand therapy, markedly accelerating scald wound healing in vivo.
12 July 2026
Read appraisal →PLoS medicine
How to benchmark medical AI agents
Medical artificial intelligence research is shifting from single-task models toward multimodal large language model-based agents for complex clinical workflows, requiring benchmarks that assess clinical reasoning, process safety, and resource stewardship rather than final outputs alone.
12 July 2026
Read appraisal →PloS one
Edge-intelligent safelink-V2X: A low-latency cooperative framework for real-time vulnerable road user protection.
The protection of Vulnerable Road Users (VRUs) remains a major challenge in modern transportation safety, as onboard line-of-sight and adverse weather conditions limit conventional onboard sensors. Existing systems that rely solely on vehicle-based sensing or on isolated communication struggle to provide timely, accurate alerts in dynamic urban environments. To address these shortcomings, this paper introduces SafeLink-V2X, a comprehensive Vehicle-to-Everything Cooperative Warning Framework designed to enhance safety for pedestrians, cyclists, and scooter riders. SafeLink-V2X employs Cellular Vehicle-to-Everything (C-V2X) and Dedicated Short-Range Communications (DSRC) protocols to enable direct data exchange of location, velocity, and heading between connected vehicles, smart infrastructure, and VRUs via smartphones or wearable tags. By applying sensor fusion and machine learning-based conflict prediction, the system identifies potential collision points and issues real-time, context-aware warnings through vehicle HMIs and VRU devices, promoting immediate evasive action. Evaluation on urban intersection simulations (detailed in Section 5) demonstrates that SafeLink-V2X reduces simulated collision probability by up to 91.4%, increases situational awareness measures by 44%, and lowers end-to-end alert latency by 30% compared to baseline onboard-only and communication-only systems under the same conditions.
12 July 2026
Read appraisal →JMIR research protocols
Service Robots as Work Support for Health Personnel in Long-Term Care: Protocol for a Scoping Review
BACKGROUND: Demographic shifts are increasing the global demand for long-term care services, coinciding with a worldwide shortage of health care personnel. Service robots, designed to perform tasks in both professional and personal use, are perceived as a potential solution to alleviate health care personnel's workload and enhance the quality of care. However, the existing literature is fragmented and heterogeneous, with a limited emphasis on the role of service robots in supporting residents rather than health care personnel. Furthermore, there is a lack of consistent definitions of service robotic technologies and a scarcity of studies on implementation models and frameworks. OBJECTIVE: This scoping review aims to map and synthesize evidence regarding the implementation of service robots as work support for health care personnel in long-term care settings. METHODS: A comprehensive 3-step search will be conducted in Embase, MEDLINE, APA PsycInfo, CENTRAL, Scopus, and CINAHL, along with gray literature databases and institutional repositories. Eligible sources encompass empirical studies and gray literature involving service robots, health care personnel, residents aged 65 years or older, and stakeholders such as informal caregivers within institutional long-term care. Exclusions apply to studies on home care, medical or industrial robots, and nonrobotic technologies. Data will be extracted and analyzed using the Joanna Briggs Institute methodology, with findings presented in tables, diagrams, and narrative summaries to identify gaps and inform future research and implementation strategies. RESULTS: The project has been funded for a 4-year period starting in April 2025. This protocol was developed in October 2025 and subsequently registered in November 2025. A comprehensive search strategy was formulated and completely conducted on October 24, 2025. The screening of 4884 titles and abstracts was completed in December 2025, resulting in the retrieval of 64 (1.3%) full-text articles for eligibility assessment. Subsequent phases, including data extraction, analysis, evidence synthesis, and presentation of results, will be conducted sequentially. The scoping review is expected to be finalized by June 2026. CONCLUSIONS: This scoping review is expected to delineate the extent and characteristics of the existing evidence on service robots as work support for health personnel in long-term care settings. It will highlight the key reported outcomes and challenges encountered in implementation studies, as well as the theoretical frameworks, models, and concepts applied to address these issues. TRIAL REGISTRATION: Open Science Framework QWK58; https://osf.io/qwk58/. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/89435.
10 July 2026
Read appraisal →European respiratory review : an official journal of the European Respiratory Society
Risk prediction for lung cancer screening: a systematic review and meta-regression.
BACKGROUND: Lung cancer (LC) remains the deadliest cancer, often diagnosed at advanced stages. Screening reduces mortality in high-risk individuals. Eligibility criteria in European and US screening guidelines have recently expanded. Therefore, we conducted an updated systematic review of risk-based models for identifying candidates for low-dose computed tomography screening and post-screening nodule classification. METHODS: We systematically searched Embase and Medline (January 2020-January 2026), identifying studies proposing new risk models in the context of LC screening. We separated models by pre- and post-screening risk stratification. Data extraction included study design, population, model type, risk horizon and model performance metrics. We performed an exploratory meta-regression of areas under the curve (AUCs) to assess whether sample size, model type, validation type and inclusion of biomarkers were associated with performance. RESULTS: Of 2462 records, 91 were included. 56 models were for screening selection (30 included biomarkers) and 35 for post-screening nodule classification. Regression-based models predominated, though machine-learning approaches were increasingly common. Discrimination ranged from moderate (AUC∼0.70) to excellent (>0.90), with biomarker and imaging-enhanced models often outperforming models without. Calibration was inconsistently reported and fewer than half underwent external validation. CONCLUSION: We identified 91 risk prediction models for LC, developed after 2020. Although many demonstrated promising discrimination across both screening selection and post-screening management, most remain insufficiently mature for clinical adoption, as their performance and practical value outside the original study setting are uncertain. Future work should prioritise external validation, updating and comparative evaluation of existing models, and prospective implementation studies rather than continued development of additional models.
10 July 2026
Read appraisal →BMC medical research methodology
Variable selection for clinical prediction models in low-dimensional data - a simulation study comparing traditional regression and machine learning methods
PURPOSE: A wide range of methods exist for developing a clinical prediction model (CPM) and for performing variable selection. Our purpose was to develop a fair simulation study design and to investigate the properties, strengths, and weaknesses of different methods to predict a continuous outcome in low-dimensional data situations. METHODS: In this simulation study, we conducted a neutral comparison of traditional (linear regression with stepwise selection) and machine learning (regularized regression with elastic net, gradient boosting, random forest) variable selection strategies to derive a CPM. The generated datasets included a total of 15 variables, with 8 of those being predictor variables. Four data- and outcome-generating mechanisms with increasing complexity produced data structures typical for biomedicine covering linear associations and gradually introducing non-linear and non-additive elements into the data structure. RESULTS: All methods generally performed better with increasing sample size and less noise in the data. Gradient boosting with regression models and with trees as base learners, and the elastic net regularized regression included nearly all variables (i.e., both the predictor and non-predictor variables), especially with increasing sample size. The linear regression model with stepwise selection (LMSS) showed the best trade-off between correctly including the predictors and excluding the non-predictor variables in most of the scenarios, even when the functional form of continuous predictors deviated from linearity. In more complex data, variable selection using the Boruta or Hapfelmeier approach for random forest performed similar to LMSS. CONCLUSION: The sample size must be sufficiently large to enable the methods to reliably identify the predictor variables and to ensure that the developed CPMs are accurate and well-calibrated. LMSS revealed good properties and the random forest with the Boruta or Hapfelmeier approach are suitable alternatives if complex associations between predictors and outcomes are assumed.
9 July 2026
Read appraisal →Journal of medical systems
Automatic Sleep Staging Using Cardiorespiratory Signals: A Systematic Review of Methodologies and Performance
Cardiorespiratory-based methods offer promising alternatives to traditional PSG for longitudinal sleep monitoring, holding significant systemic medical value for scalable sleep health management. This systematic review synthesizes methodological frameworks and performance outcomes of automatic sleep staging using cardiorespiratory signals. Four databases were searched and a total of 35 studies published since 2010 were identified. The analysis revealed that cardiorespiratory signal-based sleep staging achieved a practically meaningful accuracy of 70%, with no significant performance differences observed among signal modalities (cardiac signals, cardiorespiratory signals, or cardiac/cardiorespiratory signals combined with other non-EEG modalities) or between modeling algorithms (traditional machine learning vs. deep learning). However, we identified significant methodological heterogeneity and several critical model failure modes that hinder clinical translation, including the widespread lack of external validation, consistently poor classification of the N1 sleep stage, and limited generalization across diverse patient populations. To realize the technology's potential, future research must establish consensus-driven methodological guidelines and rigorously validate algorithms on large, demographically and clinically diverse datasets. These advances are essential for integrating cardiorespiratory-based sleep staging into healthcare systems as a scalable tool for population-level screening, longitudinal monitoring, and tiered clinical decision support.
8 July 2026
Read appraisal →Journal of biomedical informatics
A comprehensive survey of AI agents in healthcare
OBJECTIVE: This survey aims to systematically map the rapidly evolving landscape of AI agents in healthcare. It addresses the critical need to adapt general-purpose agentic frameworks characterized by autonomy, planning, and tool use to the high-stakes, safety-critical constraints of medical decision-making and patient care. METHODS: We conducted a comprehensive review of over 200 recent studies, synthesizing literature from major academic databases. We developed a holistic taxonomy that traces the full lifecycle of healthcare agents, analyzing perception modalities, core technical architectures, and evaluation protocols specific to autonomous systems. RESULTS: The review presents a quantitative landscape analysis showing exponential growth in the field. We structure the domain into three pillars: (1) Perception of multi-modal clinical data (e.g., EHR, imaging, genomics); (2) Agent Capabilities, including tool use, reasoning, memory, and multi-agent collaboration; and (3) an Application Ecosystem organized by stakeholder roles (clinicians, patients, researchers, and administrators). Additionally, we categorize evaluation frameworks, and discuss the deployment readiness of current systems across technical, evidentiary, and governance dimensions. Finally, we identify challenges for advancing healthcare agents from controlled evaluation toward real-world clinical integration. A continuously updated repository of related papers is available at https://github.com/AgenticHealthAI/Awesome-AI-Agents-for-Healthcare. CONCLUSION: AI agents offer significant potential to enhance healthcare through autonomous reasoning and workflow integration. However, current research remains largely concentrated in benchmark and controlled evaluation settings, and the translation into clinical practice will require advances in reliability, privacy protection, governance, and operational integration.
4 July 2026
Read appraisal →Journal of medical Internet research
Transformation Versus Innovation in Digital Health Care and the Future of Clinical AI
Digital innovation is frequently presented as the key to transforming health care. In this News and Perspectives article, JMIR Correspondent and academic physician Boon-How Chew reports on the distinction between and the direction of transformation and innovation, reflecting on the future of clinical AI and what lasting change in health care ultimately requires.
3 July 2026
Read appraisal →Experimental eye research
Artificial intelligence applications in OCT and OCTA for diabetic retinopathy: A systematic review
PURPOSE: Diabetic retinopathy (DR) is a leading cause of vision impairment worldwide. Optical coherence tomography (OCT) and OCT angiography (OCTA) provide detailed retinal imaging, enabling early detection of microvascular changes. This study aims to systematically review artificial intelligence (AI), particularly deep learning (DL), applications for DR detection and analysis using OCT and OCTA images. METHODS: A comprehensive literature search was conducted across PubMed, Web of Science, Scopus, IEEE Xplore, and Embase for studies published up to March 2026. A total of 1007 articles were identified, of which 175 studies met the inclusion criteria following the PRISMA study selection process. RESULTS: DL-based approaches consistently demonstrated superior performance compared to traditional machine learning (ML) methods, with reported AUC values typically ranging from 0.90 to 0.99 across classification and segmentation tasks. Convolutional neural networks (CNNs), Vision Transformers (ViTs), and encoder-decoder architectures such as U-Net showed strong performance in detecting key DR biomarkers, including microaneurysms, macular edema, and neovascularization. However, performance variability was observed depending on dataset size, imaging modality, and annotation quality. CONCLUSIONS: AI-driven analysis of OCT and OCTA images offers significant potential for automated DR detection. Despite promising results, challenges such as limited public datasets, lack of cross-institutional validation, and model interpretability remain. Future research should focus on multimodal integration, explainable AI, and large-scale validation to enhance clinical applicability.
2 July 2026
Read appraisal →International journal of medical informatics
The role of artificial intelligence in virtual emergency care: a systematic review
BACKGROUND: The integration of artificial intelligence (AI) into virtual emergency care represents a potentially transformative approach to healthcare delivery, yet the evidence base remains poorly characterized. This systematic review comprehensively evaluates the current state of AI applications in virtual emergency care settings. METHODS: We systematically searched eight databases (Embase, PsycINFO, MEDLINE, PubMed, Scopus, Web of Science, CINAHL, Cochrane Library) from inception through March 2025. Of 7,098 records identified and 4,935 screened after deduplication using Covidence, 8 studies met inclusion criteria following exclusion of one study lacking AI components. Studies were assessed using PROBAST + AI for risk of bias and quality assessment, TRIPOD + AI for reporting quality, and GRADE for certainty of evidence. RESULTS: The eight included studies (total participants: approximately 0.5 million) evaluated diverse AI applications including decision trees, machine learning ensembles, and graph neural networks across multiple virtual emergency contexts. Performance varied widely (accuracy 77.5-100%, sensitivity 63-100%, specificity 60% in single study reporting). All clinical studies demonstrated serious risk of bias. TRIPOD + AI compliance averaged only 36.9% (range 30.9-48.1%). GRADE assessment revealed very low to low certainty evidence across all outcomes, with no studies measuring actual clinical outcomes. CONCLUSIONS: Current evidence is insufficient to support widespread clinical implementation of AI in virtual emergency care. While preliminary results suggest potential benefits in triage accuracy and resource efficiency, critical gaps exist in validation, clinical outcome assessment, and reporting standards. Future research must prioritize prospective controlled trials with real patient data, clinical outcome measurements, and adherence to reporting guidelines.
2 July 2026
Read appraisal →IEEE transactions on pattern analysis and machine intelligence
A Survey on Interpretability in Visual Recognition
Visual recognition models have achieved unprecedented success in various tasks. While researchers aim to understand the underlying mechanisms of these models, the growing demand for deployment in safety-critical areas like autonomous driving and medical diagnostics has accelerated the development of eXplainable AI (XAI). Distinct from generic XAI, visual recognition XAI is positioned at the intersection of vision and language, which represent the two most fundamental human modalities and form the cornerstones of multimodal intelligence. This paper provides a systematic survey of XAI in visual recognition by establishing a multi-dimensional taxonomy from a human-centered perspective based on intent, object, presentation, and methodology. Beyond categorization, we summarize critical evaluation desiderata and metrics, conducting an extensive qualitative assessment across different categories and demonstrating quantitative benchmarks within specific dimensions. Furthermore, we explore the interpretability of Multimodal Large Language Models and practical applications, identifying emerging trends and opportunities. By synthesizing these diverse perspectives, this survey provides an insightful roadmap to inspire future research on the interpretability of visual recognition models.
1 July 2026
Read appraisal →CPT: pharmacometrics & systems pharmacology
A Narrative Review of Artificial Intelligence for Drug Repurposing: Lessons From COVID-19 and Oncology (2020-2025)
Drug repurposing presents a cost-effective and time-efficient strategy to identify new therapeutic applications for existing drugs. Recent advances in artificial intelligence, including machine learning, deep learning, knowledge graphs, and natural language processing, have revolutionized this field by enabling automated discovery of drug-disease associations. This review examines the role of artificial intelligence in drug repurposing, drawing insights from two critical case areas: Coronavirus disease 2019 and oncology. We explore current trends, methodological frameworks, and technological innovations in artificial intelligence-driven drug repurposing, as well as challenges and emerging future directions. The findings of this paper underscore the transformative potential of artificial intelligence in biomedical research and justify its continued integration in pharmaceutical pipelines.
1 July 2026
Read appraisal →ACS sensors
Lignin-Regulated Amphiphilic Janus Membrane for Eco-Friendly Electronic Skin
Eco-friendly electronic skin fabricated from sustainable natural materials represents a pivotal technology for advancing healthcare and sports-oriented wearable electronics, yet balancing comfort and electronic functionality remains challenging due to inadequate moisture management. Herein, we report an eco-friendly lignin-regulated amphiphilic Janus membrane (LAJM)-based pressure sensor that synergistically integrates unidirectional water transport with exceptional sensing performance. Through sulfonation and fluorination modifications of lignin fractions with distinct hydrophilic and hydrophobic properties, we engineer an electrospun layer with controlled hydrophobic-hydrophilic gradients, enabling unidirectional water transport while maintaining a dry and comfortable microenvironment. Moreover, LAJM demonstrates notable biocompatibility, antibacterial efficacy, and antioxidant properties, positioning it as a promising candidate for intelligent wound care systems. The LAJM-based pressure sensor achieves integrated sensing capabilities through its hierarchical structure, exhibiting a high sensitivity of 23.97 kPa-1 in the 0-20 kPa range, an extended operational range (0-200 kPa), and ultrafast response dynamics (22 ms response/31 ms recovery). These metrics enable robust continuous monitoring of vital signs (e.g., radial pulse waveforms), discrimination of voice patterns, and precise motion trajectory mapping under dynamic conditions. This technological breakthrough in lignin valorization establishes a scalable platform for next-generation breathable epidermal electronics, with transformative implications for AI-driven healthcare diagnostics and adaptive human-machine interfaces.
28 June 2026
Read appraisal →The Behavioral and brain sciences
Fair contracts for artificial intelligence?
This comment explores resource-rational contractualism (RRC) as a framework for artificial intelligence (AI) alignment and reasoning. It assesses RRC's capacity to capture the insights of philosophical contractualism and canvasses its extension to AI alignment. It suggests that a modified version of RRC, centered on impartial deliberation, can be used to identify principles for alignment and to guide efficient AI moral cognition.
27 June 2026
Read appraisal →Journal of medical Internet research
Vision-Based Artificial Intelligence Technologies for Epilepsy Monitoring: Scoping Review and Taxonomy Development Study
BACKGROUND: Artificial intelligence (AI) technologies for vision-based epilepsy monitoring are advancing rapidly in health care. Despite growing research using various video data sources and analytical approaches, no comprehensive framework exists to classify these technologies. OBJECTIVE: This scoping review aimed to develop and validate a taxonomy for AI technologies in vision-based epilepsy monitoring and to characterize visual AI approaches in epilepsy care. METHODS: Using an extended taxonomy development framework, we developed the taxonomy in 5 iterative cycles, drawing on theory and practice. We conducted a scoping review, market analysis, and applicability evaluation with market-ready solutions. We searched Scopus, Web of Science, and PubMed, including MeSH (Medical Subject Headings) terms; the final search was completed on January 16, 2026. We included primary studies from 2013 onward on AI-based or machine learning-based monitoring or prediction of epileptic seizures in humans using visual data. We excluded reviews, non-English publications, nonepilepsy studies, studies focused only on electroencephalography or wearables, animal studies, and pre-2013 publications. Evidence was charted through narrative and tabular synthesis and descriptive frequency analysis. In line with scoping review guidance, we did not conduct a meta-analysis or critical appraisal. To assess validity and practical relevance, 9 domain experts evaluated the taxonomy using a Delphi technique. RESULTS: We included 40 original studies. Study analysis yielded 16 dimensions, including data acquisition source, tracking target, image processing, classifier type, performance metrics, environment, seizure classification, data privacy, and user interface. Expert feedback added 4 further dimensions, including communication mode and information purpose. The final taxonomy comprises 23 dimensions with 102 characteristics. The review identified structural evidence gaps across settings, evaluation maturity, and reporting practices. Detection and classification in stationary settings predominated, whereas predictive approaches and real-time feedback were limited. Deep learning detection methods were common, but performance reporting was inconsistent, and patient-facing functionalities were limited. Privacy safeguards and standardized metrics were often incompletely reported, reducing comparability and maturity assessment. The taxonomy translates these patterns into guidance for benchmarking, procurement evaluation, user interface, and explainable AI design. We synthesized 5 main findings and 10 implications for research and practice. Key challenges concern standardization, seizure prediction, and real-time applicability. CONCLUSIONS: Vision-based AI technologies for epilepsy monitoring are still dominated by proof-of-concept and pilot evaluations, indicating a gap between technical feasibility and deployment-ready systems. This scoping review presents an implementation-oriented taxonomy integrating application context, system architecture, visual analysis, AI models, performance reporting, and feedback design into a single classification framework. Unlike prior work that mainly maps methods or data sources, the taxonomy provides a shared structure for consistent system-level characterization and comparison across studies and emerging solutions. It may support benchmarking, implementation-focused evaluation, procurement, and translation into clinical and home settings.
25 June 2026
Read appraisal →European journal of hospital pharmacy : science and practice
Unit dose drug dispensing systems in hospitals: a systematic review of medication error reduction and cost-effectiveness
BACKGROUND: Medical errors pose significant risks to patient safety and public health. Automated unit dose drug dispensing systems (UDDSs) have emerged as valuable tools to reduce medication errors while optimising economic and logistical resources. OBJECTIVES: This systematic review aims to evaluate studies specifically focused on the impact of automated UDDSs in reducing medication errors and streamlining processes. METHODS: A literature search was performed on PubMed, Scopus, and Web of Science, focusing on peer-reviewed articles published between 2019 and 2024. The search, concluded on 24 September 2024, included studies conducted in inpatient hospital settings that assessed automated UDDS effects on medication errors, therapy management and inventory control. Outcomes examined included effects on patient safety, cost-effectiveness and inventory management. Results were synthesised qualitatively. RESULTS: From 3346 references, four studies met the inclusion criteria: a cost-effectiveness analysis, an uncontrolled before-and-after study, and two observational studies. UDDS improved medication processes, reducing drug-related problems, medication handling and dispensing time by 50% per patient per day. Integrated with barcode scanning, UDDS lowered medication administration errors (MAEs) from 19.5% to 15.8% and harmful MAEs from 3.0% to 0.3%. Overall, medication errors dropped by 45-70%, enhancing safety and reducing manual handling risks. UDDS demonstrated cost-effectiveness by significantly reducing MAEs. The study estimated a reduction in MAEs, with a cost-effectiveness ratio of €17.69 per avoided MAE. For potentially harmful MAEs, the cost-effectiveness ratio was estimated at €30.23 per avoided error. These findings suggest substantial long-term savings potential, though the exact magnitude may vary depending on hospital size and implementation specifics CONCLUSIONS: Automated UDDSs improve patient safety by significantly reducing medication errors and delivering cost savings through better inventory management. Challenges such as high initial costs and workflow adjustments can be mitigated through gradual implementation and staff training. Further integration with other healthcare technologies, such as barcoding, real-time tracking, artificial intelligence (AI)-driven error prevention tools and fully automated restocking systems could enhance UDDS benefits and further support hospital processes.
25 June 2026
Read appraisal →JMIR medical informatics
Computational Insights Into Smart Bioelectronics in Digital Health Care (2020-2024): Topic Modeling Study
BACKGROUND: Smart bioelectronics are electronic medical devices that combine hardware and artificial intelligence (AI)-based software. These convergent medical devices analyze bio-signals measured through hardware using AI algorithms and deliver physical stimulation to enhance therapeutic effects. OBJECTIVE: This study aimed to systematically analyze recent research trends in smart bioelectronics to understand their evolving role in digital health care and to provide evidence-based insights for shaping future research and development strategies. METHODS: A total of 92 publications indexed in PubMed between 2020 and 2024 were analyzed. Latent Dirichlet allocation-based topic modeling, optimized using coherence scores, was applied to identify latent research themes. RESULTS: The results indicate a steady increase in related research over the past 5 years, along with a clear shift in research focus from bio-signal sensing and bioelectronic device materials toward AI-driven analysis and disease-oriented applications, ultimately evolving into intelligent and adaptive bioelectronic therapeutic systems. Three major research topics were identified: bio-signal-based neuromodulation (n=23, 25%), AI-driven neurological disease analysis (n=32, 34.7%), and implantable bioelectronics and biomaterials (n=37, 40.2%). CONCLUSIONS: By mapping the evolving landscape of smart bioelectronics, this study provides valuable insights into their multidisciplinary development and highlights their potential applications in clinical decision support, personalized rehabilitation, and next-generation medical device innovation.
24 June 2026
Read appraisal →International ophthalmology clinics
Beyond the Clinic: Home-Based Monitoring Strategies in Glaucoma Care
Glaucoma is a chronic and progressive optic neuropathy, necessitating lifelong monitoring for affected individuals. Traditional surveillance of the disease relies on evaluation of intraocular pressure (IOP), optic nerve anatomy, and visual function at the physician's office, which only provides limited snapshots of a disease characterized by dynamic physiological variation. Developments in portable ophthalmological technologies and digital health platforms have enabled home monitoring of glaucoma. This review synthesizes current evidence on the home monitoring of glaucoma, with a focus on self-tonometry and home-based visual field testing. Handheld tonometers such as iCare HOME and iCare HOME2 allow repeated, anesthesia-free self-measurement of IOP in a real-world setting. These devices have demonstrated good agreement with Goldmann applanation tonometry, the gold standard. Home tonometry captures diurnal variations of IOP, thus providing relevant insights for diagnosis and therapeutic decision-making. Home-based visual field testing (eg, tablet-based platforms, virtual reality-based testing) enables frequent, unsupervised functional assessment with repeatability and sensitivity comparable to standard automated perimetry. Increased testing frequency may facilitate earlier detection of rapid progression, particularly in the early stages of disease. Collectively, evidence suggests the feasibility and clinical relevance of these home monitoring modalities. However, patient compliance with testing schedules and cost-effectiveness remain as challenges.
24 June 2026
Read appraisal →Journal of robotic surgery
Emerging robotic platforms in gynecologic surgery: a systematic review.
Robotic innovation in gynecologic surgery has expanded beyond established multiport ecosystems toward single-port, modular multi-arm, transvaginal, AI-enabled, and teleoperated platforms. However, available evidence remains fragmented and unsuitable for quantitative synthesis. This systematic review evaluated emerging robotic platforms in gynecologic surgery, focusing on procedure types, outcomes, costs, and learning curves. The review was prospectively registered in PROSPERO (CRD420261360389). Eligible studies included clinical studies evaluating emerging robotic platforms in gynecologic surgery. PubMed, Scopus, and Web of Science were searched on 4 April 2026. After duplicate removal, 408 records were screened, 160 full texts assessed, and 92 studies included. Data extraction and risk of bias assessment were performed independently in duplicate. The 92 studies covered 13 platforms and were mainly retrospective single-center studies. Evidence was most extensive for da Vinci SP and HUGO RAS, with more limited but informative data for Senhance, Versius, hinotori, and KANGDUO. Evidence for other systems consisted mainly of first-in-human, pilot, or feasibility studies. Hysterectomy and sacrocolpopexy were the most frequently reported procedures. Overall, perioperative outcomes and short-term safety appeared favorable, despite limited follow-up and heterogeneous outcome definitions. Cost data were mainly available for Senhance and KANGDUO. Learning-curve evidence was strongest for da Vinci SP and HUGO RAS. Risk-of-bias was predominantly moderate. Emerging robotic platforms have demonstrated broad technical feasibility and generally acceptable short-term perioperative safety in selected gynecologic procedures, but current evidence is heterogeneous, observational, and methodologically limited. Rigorous prospective studies, standardized economic analyses, longer follow-up, and learning-curve evaluations are needed before comparative conclusions can be drawn.
23 June 2026
Read appraisal →Journal of robotic surgery
Telesurgery and remote surgery research: a bibliometric analysis of scientific growth, collaboration, and thematic evolution
Telesurgery and remote surgery have gained relevance in surgical innovation due to advances in robotics, telepresence, artificial intelligence, and high-speed communication networks. Despite this progress, the global evolution, collaborative structure, and thematic development of this field remain insufficiently characterized. A bibliometric study was conducted using records retrieved from Web of Science. Publications on telesurgery and remote surgery published up to 2025 were included. Bibliographic data were processed and analyzed using the bibliometrix package in R. Bibliometric indicators, citation metrics, country productivity, journal output, author contributions, collaboration networks, and keyword-based thematic patterns were assessed. A total of 476 publications on telesurgery were included, distributed across 290 journals. Scientific production showed sustained growth, with an annual growth rate of 20.59%, a mean document age of 6.06 years, and peak output in 2024. The average citation count was 28.54 citations per document. The United States and China were the leading contributors, with 151 and 120 publications, respectively, and also acted as the main hubs in the international collaboration network. The most cited studies focused on telerobotic systems for minimally invasive surgery, telerobotic neurovascular interventions, and ultra-reliable low-latency communications. Keyword analysis showed a progressive shift from early clinical feasibility and surgical applications toward connectivity, latency, robotic systems, extended reality, blockchain, and artificial intelligence. These trends indicate the transition of telesurgery from a clinically exploratory field to a technology-driven research domain. Telesurgery and remote surgery constitute a rapidly expanding and technologically driven research field. Future progress will require stronger international collaboration, real-world clinical validation, standardized outcome assessment, and strategies to reduce technological inequities across regions.
23 June 2026
Read appraisal →Journal of robotic surgery
Mapping the scientific landscape of robotic hernia repair: a bibliometric and topic modeling analysis of thematic transitions
Robotic hernia repair (RHR) has profoundly transformed modern surgery, yet it remains intensely debated. To understand this dynamic field, bibliometric analysis offers a powerful tool to evaluate its evolving research trends. Our study presents a comprehensive bibliometric and machine learning analysis of 1,503 publications from 2003 to 2025. The literature demonstrates a growth trend (R2 = 0.96), characterized by initial latency (2003-2015), steady adoption (2016-2020) and a recent exponential explosion (2020-2025) peaking at 262 publications in 2025. The United States is the leading contributing nation (n = 892), followed by Italy (n = 98), Germany (n = 71), and China (n = 68). Key contributors include Kudsi OY, Malcher F, and Muysoms F, with Surgical Endoscopy serving as the predominant publication venue (n = 203). Spatial and relational networks were mapped using VOSviewer and CiteSpace. Critically, Latent Dirichlet Allocation (LDA) topic modeling combined with principal component analysis (PCA) stratified the corpus into four major research domains comprising the anatomical spectrum, the balance between resource and risk management, hardware evolution and surgical education, and technical complexity coupled with high-level evidence. Longitudinal regression of thematic probabilities revealed a significant thematic transition in research emphasis from early hardware feasibility and technical exploration toward value-based evaluation. Current research priorities are increasingly focused on rigorous cost-effectiveness evaluation in high-risk populations together with the generation of high-quality randomized evidence. In conclusion, this multidimensional analysis illustrates how this field is increasingly guided by the interconnected domains of Evidence, Education and Economics, which align with established frameworks for sustainable surgical progress.
21 June 2026
Read appraisal →Journal of robotic surgery
Rethinking health technology assessment in robotic surgery: an EFISDS-TROGSS position paper. Official position paper of the European Federation - International Society for Digestive Surgery (EFISDS) and The Robotic Global Surgical Society (TROGSS).
Robotic-assisted surgery (RAS) has evolved from a procedural innovation into an increasingly integrated component of contemporary digital surgical ecosystems. Nevertheless, most current Health Technology Assessment (HTA) frameworks continue to evaluate robotic systems primarily through comparator-based models focused on isolated perioperative and oncological outcomes. In this EFISDS-TROGSS position paper, we critically examine the methodological limitations of conventional HTA paradigms when applied to robotic surgical platforms, using the recent Italian AGENAS appraisal as a representative case study. While the AGENAS document represents one of the most comprehensive national evaluations of RAS performed to date, its heterogeneous recommendations across procedures highlight unresolved tensions regarding perioperative benefit, real-world implementation, learning curves, organizational impact, and long-term healthcare value. We argue that RAS should increasingly be interpreted not simply as a surgical device, but as a platform technology interacting with simulation-based training, digital infrastructure, surgical data science, artificial intelligence, telecommunication systems, and institutional organization. Rather than a surgical device alone, RAS should be interpreted and regarded as a combination of technological advances and approaches that integrate various degrees of artificial intelligence autonomy, image navigation, telesurgery, and other benefits to empower the surgical team. Conventional HTA models, originally developed for relatively discrete therapeutic interventions, may incompletely capture the multidimensional interaction between robotic technologies and modern healthcare systems. Particular attention is dedicated to real-world evidence, implementation maturity, reimbursement limitations, and the growing mismatch between current Diagnosis-Related Group (DRG) structures and technologically integrated surgical care. Finally, we propose more flexible and multidimensional assessment frameworks integrating procedural outcomes with organizational sustainability, digital interoperability, workforce implications, and longitudinal healthcare value.
20 June 2026
Read appraisal →International journal of medical informatics
Human in the loop artificial intelligence in healthcare: applications, outcomes, and implementation challenges
BACKGROUND: The integration of artificial intelligence in healthcare has transformed clinical practice and research methodologies. However, concerns regarding algorithmic accountability, interpretability, and safety have necessitated human oversight in AI systems. Human in the loop artificial intelligence represents a collaborative paradigm where human expertise and machine intelligence converge to enhance decision making while maintaining ethical standards and clinical safety. AIM: This review synthesizes current evidence on human in the loop AI in healthcare delivery and research, examining implementation frameworks, clinical outcomes, comparative advantages over fully automated and clinician-only approaches, and challenges. METHOD: A comprehensive narrative review was conducted using PubMed, Scopus, Web of Science, and IEEE Xplore databases covering studies from 2018 to 2025. Data were thematically synthesized to identify patterns, frameworks, and outcomes. This narrative approach enables comprehensive conceptual synthesis across diverse HITL-AI applications and contexts. RESULTS: Human in the loop AI demonstrates significant applications across diagnostic imaging, clinical decision support, patient monitoring, drug discovery, and research data analysis. Evidence indicates improved diagnostic accuracy, reduced medical errors, enhanced patient safety, and increased clinician trust compared to both automated AI and traditional approaches. Implementation requires EHR interoperability, clear liability frameworks, adaptive training protocols, and quantum-safe cryptographic security. Challenges include workflow integration, regulatory gaps for adaptive systems, and sustainability concerns. CONCLUSION: This review advances the field by synthesizing cross-domain implementation patterns, mapping collaboration models to risk-stratified contexts, identifying regulatory gaps for adaptive systems, and proposing future directions including post-quantum cryptographic integration, AI-driven adaptive architectures, and multi-center scalability frameworks for optimizing human-machine collaboration in healthcare.
15 June 2026
Read appraisal →Journal of medical systems
MEL-IA: An Interoperable AI System for Multimodal Skin Lesion Classification in Hospital Settings
UNLABELLED: Skin cancer represents a growing healthcare challenge, motivating the development of AI sytems that can support lesion assessment while integrating efficiently into clinical workflows. This study presents MEL-IA (MobilE skin Lesion dIAgnosis), an interoperable AI system designed for automated skin lesion classification and full integration with hospital information systems. METHODS: The system combines dermatoscopic images and structured clinical metadata using an EfficientNet-B4 - based multimodal model trained on the ISIC 2019 BCN_20000 and MSK subsets. Internal robustness was assessed through stratified 5-fold cross-validation, and external generalization was evaluated on the independent HAM10000 dataset. Interoperability and operational performance were validated through deployment in a real hospital environment using HL7, DICOM, PACS, and HIS/RIS systems. RESULTS: We evaluated internal robustness using stratified five‑fold cross‑validation, with global accuracy of 0.86, macro F1-score of 0.85, and AUC values above 0.97. The external evaluation showed that the model generalizes well under different acquisition conditions, achieving an accuracy of 0.84 and balanced accuracy of 0.65, with sensitivities of 0.60 for melanoma 0.72 for basal cell carcinoma. The multimodal model outperformed the image model baseline across all classes. Technical deployment confirmed full interoperability, > 99% successful study integration, and near real time processing. CONCLUSIONS: MEL-IA provides multimodal skin lesion classification capabilities and integrates smoothly into hospital infrastructures based on standards. The results demonstrate technical feasibility, interoperability, and operational viability for deployment within dermatology oriented clinical workflows. Further clinical validation studies are warranted to assess the system's impact on diagnostic decision-making and patient outcomes.
14 June 2026
Read appraisal →Journal of medical Internet research
Patient Cognitive Bias in Large Language Model-Supported Health Consultations: Simulation-Based Comparative Study
BACKGROUND: Large language models (LLMs) are increasingly used by patients for health information and preliminary medical advice. In patient-facing consultations, users may present explicitly stated diagnostic preferences or symptom narratives emphasizing a preferred explanation. Such cognitively biased input constrains the diagnostic context available to the model and may systematically steer its reasoning during interactive LLM-supported health consultations. OBJECTIVE: This study aimed to quantify the impact of patient cognitive bias on LLM diagnostic performance in multiturn consultations, assess the effectiveness of prompt-based mitigation strategies and decoding temperature adjustment, and evaluate a dual-system framework for improving robustness under biased interaction. METHODS: We developed a simulated patient agent to generate both unbiased and cognitively biased consultations using 1273 medical question answering dataset United States Medical Licensing Examination cases. Six widely used LLMs of varying capacities were evaluated through 3-round, multiturn dialogues, after which each model produced a final diagnostic judgment based on the complete consultation record. Diagnostic accuracy was the primary outcome. Secondary outcomes included bias-induced accuracy decline (absolute reduction in accuracy under biased vs standard consultations) and bias-influenced error proportion (proportion of incorrect responses aligned with the patient's preferred but incorrect diagnosis). Three prompt-based mitigation strategies and 4 decoding temperature settings were tested. In addition, a dual-system framework was evaluated, in which a conversational foundation LLM conducted patient interaction and history taking (System 1), while a reasoning-oriented LLM (o1-mini) generated the final diagnostic judgment (System 2). In the foundation-only condition, the same LLM performed both interaction and diagnosis. RESULTS: Across all 6 evaluated models, cognitively biased consultations led to marked diagnostic accuracy declines of approximately 7 to 39 percentage points compared with standard multiturn consultations, whereas static single-response tests and standard consultations showed comparable accuracy. Larger deteriorations were observed in lower-capacity models, with some approaching random-guess performance under bias. Errors were frequently aligned with patient bias, with bias-influenced error proportion exceeding one-third across models, indicating systematic conformity rather than random error. Prompt-based mitigation strategies and decoding temperature reduction yielded limited and inconsistent improvements and did not reliably prevent bias-induced performance loss. By contrast, the dual-system framework substantially improved diagnostic accuracy under biased conditions, producing gains of approximately 10 to 39 percentage points across most models and recovering a large proportion of the performance lost due to bias, particularly in lower-capacity systems. CONCLUSIONS: Patient-driven cognitive bias represents an underrecognized behavioral risk in LLM-supported health consultations. Common mitigation approaches, such as prompt engineering or decoding parameter adjustment, provide limited resilience. Explicitly separating conversational interaction from deliberative diagnostic reasoning through a dual-system framework enables more robust diagnostic performance under biased input while potentially preserving patient-facing dialogue fluency by retaining the foundation LLM as the conversational component, offering a scalable design strategy for safer medical AI systems.
14 June 2026
Read appraisal →Scientific reports
AIDx: a locally deployable AI system for physician clinical decision support
The dynamic environment of medicine, particularly in settings such as the Emergency Department, challenges physicians with an influx of patient data and the need for swift diagnosis and treatment under high-pressure conditions. Numerous AI models have been developed for medical applications, yet their integration into hospital systems remains limited because they often do not align with existing clinical workflows. To address these challenges, I developed AIDx, an AI-powered system designed to help physicians streamline clinical decision making, improve diagnostic support, and provide an integrated platform for AI-assisted analysis. Unlike many standalone systems, AIDx is designed for interoperability with electronic health record (EHR) systems and can leverage supplementary tools such as updated medical knowledge bases and auxiliary AI models. At its core, AIDx includes AIDx-Copilot, a large language model fine-tuned on de-identified EHR data and optionally grounded using retrieval-augmented generation (RAG) from open medical references. This study is a text-only benchmark assessment. I evaluated AIDx-Copilot using the MultiMedQA benchmark suite under a unified, deterministic, single-pass protocol (temperature=0; no voting). Across nine MultiMedQA subsets (MedQA, PubMedQA, MedMCQA, and medically focused MMLU subjects), AIDx-Copilot achieves a mean accuracy of 83.61% (SD 7.37). I report per-dataset Wilson 95% confidence intervals to quantify uncertainty from finite test set sizes. To isolate component contributions, I conducted ablation experiments comparing the base model, the fine-tuned model without retrieval, and the fine-tuned model with RAG enabled. EHR-based fine-tuning accounts for the majority of the performance gain (+17.8 percentage points over the base model on average), while RAG provides a modest additional benefit (+0.4 points on average) that varies by dataset. A qualitative error analysis of 200 incorrectly answered items identifies knowledge gaps (41.0%) and reasoning errors (38.0%) as the dominant failure modes. Comparative numbers for larger proprietary systems are provided only as context in the supplementary material because they may use different prompts and inference settings. The deployment configuration (quantized weights and on-premises serving) supports fast inference and local deployment. Under the quantized configuration, the system achieves a median latency of 0.84 seconds per query and fits within 28.1 GB of VRAM across two commodity GPUs. The primary results are based on public benchmark evaluations without RAG; I did not perform clinical, user-study, or real-world validations. In addition, while AIDx is designed to support fully local operation, specific deployments may choose to use third-party embedding or vector-search services; this paper documents the data-flow boundary and a fully local configuration path.
14 June 2026
Read appraisal →PloS one
Gauging support for autonomous delivery drones powered by artificial intelligence: A public survey in Singapore
There has been emerging empirical evidence supporting the role of artificial intelligence (AI) knowledge in public support for AI. These findings stand in sharp contrast to the long-standing cognitive miser view in determining technological acceptance. In this study, by contrasting models of knowledge deficit against cognitive miser using data from a large-scale public opinion survey about autonomous delivery drones in Singapore (N = 1,002), we investigated whether knowledge and various heuristic cues were associated with AI support. Our results revealed that objective knowledge was not a direct determinant of support, while various heuristic cues such as propensity to trust in AI, news media attention, perceived source credibility in scientists and authorities, as well as risk and benefit perceptions presented significant associations. Mediation analyses further revealed knowledge was indirectly associated with support through heuristic cues such as source credibility perceptions as well as risk and benefit perceptions. Our findings were largely consistent with the cognitive miser model in that AI acceptance is immediately associated with cognitive heuristics rather than knowledge. Theoretical and practical implications are discussed.
12 June 2026
Read appraisal →Journal of robotic surgery
Technical challenges in autonomous robotic ultrasound examinations: perception, planning, and control
Ultrasound is one of the most widely used medical imaging modalities in clinical practice. Robot-assisted ultrasound systems (RAUS) offer significant advantages in reducing physician workload and improving imaging standardization. However, the dynamic and complex nature of the human physiological environment severely hinders the transition of RAUS from automation to full autonomy. This paper presents a narrative review of the core research progress of RAUS between 2021 and early 2026. It comprehensively analyzes the current technological landscape of autonomous robotic ultrasound by focusing on three core pillars: perception, planning, and control. By evaluating cutting-edge breakthroughs and existing limitations across these three domains, this paper identifies the key technical bottlenecks that restrict fully autonomous RAUS. Finally, we propose an evolutionary roadmap driven by multimodal fusion, digital twins, planning-control co-design, and embodied artificial intelligence (AI) architectures. This roadmap provides a developmental reference for autonomous robotic ultrasound examinations and also lays a crucial foundation for future autonomous ultrasound-guided surgical interventions.
1 June 2026
Read appraisal →Journal of medical Internet research
Acceptability of Technologies to Support Early Dementia Detection: Qualitative Study With the Boston University Alzheimer's Disease Center Cohort
BACKGROUND: Dementia is on the rise globally due to increasing life expectancies and population growth. Digital technologies may help detect early signs, enabling timely interventions to slow or reverse cognitive decline. However, to support the successful implementation of these digital technologies into health care settings, they must be acceptable to target users. Older adults and those with mild cognitive impairment (MCI) are at risk of developing dementia in later life and need to be able to use these technologies in order for this intervention to be approved and implemented in clinical practice. OBJECTIVE: This study explored the perspectives of older adults and those living with a clinical diagnosis of MCI on the acceptability of using various digital technologies that have the potential to support early dementia detection. METHODS: Participants were recruited from Boston University's Alzheimer's Disease Research Center. Participants selected at least 2 technologies from 9 different wearables and software to use for 2 weeks, at 3-month intervals, over a total duration of 2 years. A subgroup of self-selecting participants was interviewed after the first 2 weeks of use to gather initial perspectives regarding the acceptability of using the digital technologies. An inductive framework thematic analysis approach was used, assisted by NVivo (version 14.23.2; QSR International). RESULTS: In total, 13 individuals living with a clinical diagnosis of MCI and 11 adults aged 65 years and older were interviewed. Our analysis identified five key themes: (1) gamification, (2) wearability, (3) user guidance, (4) burden of use, and (5) usefulness. Gamified apps were generally liked, although users with little experience of digital games needed time to adjust. Wearables resembling everyday accessories (eg, watches) were preferred, but complaints about tight or uncomfortable straps were frequently reported. Clear instructions were critical to support correct use, but many participants would have liked more troubleshooting support when technical issues arose. The use of 5 or more devices led to a high burden, especially when devices had practicality issues such as not being waterproof. Devices offering personal feedback were perceived as useful to satisfy personal interests, though some questioned their usefulness within health care. Participants raised concerns about losing valued personal interactions with health care professionals and questioned how their existing health conditions and treatment for such conditions may affect the validity of the data collected by the devices. CONCLUSIONS: These findings can guide researchers in choosing appropriate devices and minimizing burden. Future work should explore the views of those experiencing digital exclusion to ensure equitable access to dementia-detection technologies.
31 May 2026
Read appraisal →Journal of medical Internet research
Biocomputing: Beyond the Hype
Biocomputing is a nascent but rapidly developing field at the intersection of biology and computer science. In this News and Perspectives article, JMIR Correspondent Simon Spichak reports on its current and potential applications for health care research and beyond.
29 May 2026
Read appraisal →Philosophical transactions. Series A, Mathematical, physical, and engineering sciences
Reproducible workflow for online artificial intelligence in digital health
Online artificial intelligence (AI) algorithms are an important component of digital health interventions. These online algorithms are designed to continually learn and improve their performance as streaming data are collected on individuals. Deploying online AI presents a key challenge: balancing adaptability of online AI with reproducibility. Online AI in digital interventions is a rapidly evolving area, driven by advances in algorithms, sensors, software and devices. Digital health intervention development and deployment is a continuous process, where implementation-including the AI decision-making algorithm-is interspersed with cycles of re-development and optimization. Each deployment informs the next, making iterative deployment a defining characteristic of this field. This iterative nature underscores the importance of reproducibility: data collected across deployments must be accurately stored to have scientific utility, algorithm behaviour must be auditable and results must be comparable over time to facilitate scientific discovery and trustworthy refinement. This article proposes a reproducible scientific workflow for developing, deploying and analysing online AI decision-making algorithms in digital health interventions. Grounded in practical experience from multiple real-world deployments, this workflow addresses key challenges to reproducibility across all phases of the online AI algorithm development life cycle. This article is part of the theme issue 'Statistical workflow'.
29 May 2026
Read appraisal →Journal of robotic surgery
Global evolution of robot-assisted cholecystectomy research in the era of artificial intelligence: a bibliometric and knowledge-mapping study
OBJECTIVE: To systematically map the global evolution, collaborative networks, knowledge structure, and emerging research hotspots in robot-assisted cholecystectomy (RAC) using bibliometric and visualization techniques. METHODS: A comprehensive bibliometric analysis was conducted using the Web of Science Core Collection (2005-2025). Publications were retrieved using predefined search strategies and screened according to strict inclusion criteria. CiteSpace, VOSviewer, and R software were employed to analyze publication trends, co-authorship networks, institutional and national collaborations, co-citation patterns, keyword co-occurrence, and research bursts. Knowledge mapping techniques were used to visualize thematic evolution and intellectual structure. RESULTS: A total of 926 eligible publications were included. Global output demonstrated a continuous and marked increase, particularly after 2016, with citations following a similar upward trajectory, reflecting growing academic impact. The United States dominated both publication output and citation influence, while collaboration networks remained largely regionally clustered with limited cross-national integration. Research was primarily concentrated in high-impact surgical journals, with foundational contributions emphasizing feasibility, safety, and comparative effectiveness. Keyword and co-citation analyses revealed a knowledge structure centered on laparoscopic surgery, bile duct injury, and robotic systems, with emerging clusters highlighting artificial intelligence (AI) and surgical education. Evolutionary trajectory analysis demonstrated a transition from technical feasibility (early stage), to safety and evidence-based evaluation (middle stage), and to training and technological integration (recent stage). Burst detection further identified recent hotspots in AI, cost-effectiveness, and surgical training systems. Emerging evidence also indicates increasing integration of AI-driven systems into RAC, enabling intraoperative decision support, workflow recognition, and early-stage semi-autonomous surgical execution, particularly in standardized procedures such as cholecystectomy. CONCLUSION: RAC research has evolved from early exploratory studies toward increasing technological integration and methodological refinement. Although RAC is not currently among the most dominant clinical indications for robotic surgery, it provides a valuable model for studying surgical standardization, training systems, and emerging intelligent surgical technologies. While current applications remain largely assistive, emerging advances in AI and robotic technologies suggest the potential for future development toward more intelligent and partially automated surgical systems. However, most evidence currently remains experimental and has not yet translated into widespread clinical practice. Future research should focus on strengthening global collaboration, improving high-quality evidence generation, and carefully evaluating the safety, feasibility, and clinical applicability of emerging intelligent surgical technologies.
26 May 2026
Read appraisal →Journal of medical Internet research
Patients' Perspectives on the Implementation of AI in Radiological Diagnostics: Focus Group Study
BACKGROUND: Rapid developments in artificial intelligence (AI) will enable its widespread use in radiological diagnostics in the near future. Patients will then be confronted with findings generated with the help of AI. Understanding patients' perspectives on the use of this technology is one of the key factors for its successful implementation. OBJECTIVE: This qualitative study aimed to gain insight into patients' reasoning about the opportunities and risks of using AI in radiological diagnosis, to identify prerequisites for its acceptance, and to identify aspects that can promote trust in AI diagnoses, especially in scenarios of high personal concern. METHODS: A total of 7 focus groups were conducted with 34 patients (n=15, 44% female participants) aged between 23 and 85 years (mean 49.06, SD 17.08 y), recruited using purposive sampling strategies. Each focus group was audiotaped, transcribed, and analyzed using the method of structured qualitative content analysis. RESULTS: Study findings show that patients are open to the use of AI in radiological diagnostics. The basic prerequisites for this are (1) scientific evidence of safe outcomes that are more accurate and faster than those without AI; (2) recognizable added value in patient care; (3) transparency in the use of AI and disclosure to the patient; (4) comprehensive, binding measures for quality assurance; and (5) the use of AI solely to support the physician. However, the results indicate that further criteria are important for patients to be willing to choose a radiologist who uses AI and to trust AI diagnoses. In situations where they are personally affected, patients fear that physicians will place too much trust in the AI result and that the physician-patient relationship will become dehumanized. Therefore, the physicians' abilities and functions that inspire trust from the patients' perspective must come into play. These include (1) an independent diagnosis by the physician that takes into account not only the clinical context but also the individuality of the patient, (2) a comprehensible explanation of the pros and cons of using AI for patients and clear communication of AI output, and (3) a humane and empathetic physician-patient relationship, which shows that the physician continues to feel responsible for the patient. CONCLUSIONS: The results of the study underscore that a high quality of the entire "sociotechnical" system is an essential prerequisite for patient acceptance of the use of AI in radiological diagnostics and for trust in AI diagnoses. The further development of AI performance must go hand in hand with the creation of framework conditions for its use that meet patients' expectations of the role of the physician and ensure a trust-building physician-patient relationship. The study provides valuable insights into how such integration of AI into radiological practice can be achieved.
26 May 2026
Read appraisal →JMIR nursing
Health Care Professionals' Perspectives on the Use of a Wearable Device for Early Detection and Continuous Vital Signs Monitoring of Acute Respiratory Infections in Nursing Homes: Qualitative Study
BACKGROUND: The growing aging population and staff shortages are placing pressure on Dutch nursing homes (NHs). These challenges have led to an increased interest in digital health technologies. Among these are wearable devices that allow for remote continuous monitoring of vital signs. An example is the Healthdot (smartQare), a wearable electronic device that continuously monitors heart rate, respiratory rate, and physical activity. In the context of acute respiratory infections (ARIs) in NHs, where initial symptoms can go unnoticed, continuous monitoring may aid in early recognition, timely intervention, and reduce staff workloads. However, little is known about how health care professionals perceive the use of continuous vital signs monitoring devices, such as the Healthdot, for this cause in NHs. OBJECTIVE: This study aims to explore the perspectives of healthcare professionals on the use of the Healthdot for early detection and monitoring of ARIs in NHs, to inform potential future implementation. METHODS: Semistructured interviews were conducted with 20 physicians, nurses, and certified nursing assistants from 4 NHs and 1 acute geriatric community hospital located in a NH. Interview transcripts were thematically analyzed to identify themes regarding their perspectives on the use of the Healthdot for monitoring ARIs in this setting. RESULTS: Five main themes were identified that related to the appropriate use of the Healthdot for NH clients and health care professionals: alignment of Healthdot use and NH clients' treatment policies, balancing safety and freedom, impact of the Healthdot on work processes, supporting rather than replacing care, and possible use during pandemics and in the future. Additionally, several preconditions for the use of the Healthdot were identified, including its usability, a support base among care staff, adequate training and guidance, communication with NH clients and their relatives, and a clear policy regarding its use. CONCLUSIONS: Given the complexity of care in NHs, where clinical care is typically balanced against quality of life and a homelike environment, physicians generally expressed reserved attitudes toward the Healthdot, highlighting the need to consider multiple factors in its implementation. Care staff were generally positive about the device. Nevertheless, tailored assessment for each individual NH client remains essential, balancing treatment goals, safety, autonomy, and person-centered care. Additionally, clear communication and alignment between health care professionals in this setting are crucial, specifically regarding their expectations of the Healthdot's role in care processes. This study offers practical guidance that may inform future implementation efforts of continuous vital sign monitoring devices in NHs.
26 May 2026
Read appraisal →BMJ health & care informatics
Beyond the digital health tsunami: reaffirming biomedical and health informatics as the scientific foundation of digital healthcare transformation
Digital health (DH) has emerged as the dominant terminology for technology-enabled healthcare, reshaping institutional discourse, educational programmes and policy agendas. DH is often portrayed as a novel discipline that could supersede Biomedical and Health Informatics (BMHI), creating conceptual ambiguity and the risk of weakening scientific rigour in the design, evaluation and governance of digital interventions. This paper clarifies BMHI's disciplinary identity, characterises DH as a practice-oriented and implementation-oriented domain, and establishes the epistemic relationship between the two in order to inform academic strategy, curricula and professional identity. A comparative conceptual analysis examined scope, theoretical foundations, methodological approaches, educational frameworks, professional structures and stakeholder ecosystems in both domains, informed by recent debates on degree programmes and workforce training. We found that BMHI is a mature scientific discipline with explicit theoretical foundations encompassing data representation, knowledge modelling, clinical reasoning, decision support, interoperability, sociotechnical analysis, evaluation science, safety, ethics and governance. DH, by contrast, represents an implementation-oriented translational domain that operationalises BMHI through products, services and organisational transformation (eg, telehealth, mobile applications, artificial intelligence-enabled clinical services). While DH has amplified visibility and accelerated adoption, it lacks autonomous theoretical foundations and depends on BMHI for scientific validity, safety and sustainability. Three evolutionary scenarios are examined: DH absorbing BMHI (epistemically incoherent), convergence as parallel disciplines (conceptually flawed), and reintegration of DH within BMHI as its applied layer (scientifically coherent and strategically sustainable) but requiring disciplinary communication, curricular alignment and institutional leadership explicitly positioning DH as a translational layer of BMHI.
24 May 2026
Read appraisal →Studies in health technology and informatics
Mapping Existing Evidence on Physicians' and Patients' Experiences with GenAI in Clinical Communication and Documentation: A Rapid Review
GenAI tools like chatbots and ambient scribes are increasingly used in clinical documentation, raising ethical and safety concerns. This rapid review maps existing evidence on experiences and attitudes of physicians and patients toward using GenAI in clinical communication and documentation. Structured searches were conducted in PubMed and Web of Science. Studies were screened using PICO criteria. Data were extracted and summarized descriptively, with high-order thematic analysis of research aims and emerging recommendations. 41 studies were included: 75.6% focused on physicians, 9.8% examined patient perspectives exclusively, and 14.6% included both. Some of the themes identified were usability, exploration of perceptions and expectations, acceptability, and ethical implications. Recommendations emphasized the need for diverse cohorts, real-world validation, and ethical considerations. Current literature was skewed toward physician perspectives. Broader, meaningful inclusion of patients is needed to support equitable and trustworthy GenAI integration in healthcare.
24 May 2026
Read appraisal →Studies in health technology and informatics
Towards Clinical NLP in a Low-Resource Language: Structuring Estonian Clinical Text
Healthcare sectors generate large volumes of unstructured text, yet mature clinical NLP is lacking for low-resource languages. This project targets Estonian, a morphologically rich, low-resource language, and outlines an end-to-end pipeline that converts free text into interoperable, structured data. A design-science approach guides development of multilingual speech-to-text for clinical dictation, clinical NLP for entity extraction and context handling, terminology binding to the Estonian SNOMED CT extension, and export as HL7 FHIR resources with mappings to EHR fields. Expected benefits include reduced documentation time through dictation with automatic structuring, improved data quality and consistency via controlled terminology, and faster retrieval of clinically relevant information to support decision-making. Pilot evaluations with healthcare partners will assess accuracy, usability, and workflow impact.
24 May 2026
Read appraisal →Studies in health technology and informatics
Frameworks for Digital Health Engagement: A Scoping Review Protocol
Digital health technologies (DHTs) are increasingly integral to healthcare delivery, yet sustained engagement remains inconsistent across users and settings. This scoping review aims to map frameworks and instruments that conceptualise digital health engagement across adoption and sustained use. Searches were conducted in Scopus, Web of Science, ACM Digital Library, and PsycINFO. The review will clarify conceptual coverage and identify gaps in digital engagement.
24 May 2026
Read appraisal →Studies in health technology and informatics
Logic Models on Health Information Technology: First Insights from a Cross-Disciplinary Scoping Review
BACKGROUND: Health information technology (HIT) interventions are complex, yet theory-based approaches such as logic models or Theory of Change (ToC) remain rarely applied in medical informatics, although they could clarify assumptions, link activities to outcomes, and support design and evaluation. OBJECTIVES: This scoping review examined how logic models and ToC have been used in HIT research across disciplines. METHODS: Following the PRISMA-ScR guidelines, databases were searched up to August 2025. Eligible studies explicitly applied logic models or ToC to HIT. Data were extracted on HIT type, purpose of using logic models, and reported benefits and challenges. RESULTS: Sixty-nine studies met inclusion criteria, most published after 2020. A majority of studies developed logic models for patient-facing HIT such as mHealth or telehealth. Logic models were mainly used to support HIT design or evaluation and were mostly newly developed. Reported benefits of using logic models included conceptual clarity and better stakeholder communication; challenges involved resource demands and limited generalisability. CONCLUSIONS: Logic models and ToC are increasingly used for HIT but remain inconsistently applied and seldom reused. Standardised reporting and repositories could strengthen theory-informed evaluation in medical informatics.
24 May 2026
Read appraisal →JMIR human factors
Development and Psychometric Validation of a Usability Instrument Based on ISO 25010 for Electronic Health Record Systems in Peruvian Health Care Settings: Methodological Study
BACKGROUND: The usability of electronic health records influences patient safety, clinical efficiency, and the sustainable digital transformation of health care. While tools such as the System Usability Scale exist, few are based on structured software quality models or align with local regulatory frameworks, particularly in low- and middle-income countries. OBJECTIVE: This study aims to develop and psychometrically validate an instrument to assess electronic health record usability, based on the ISO 25010 (International Organization for Standardization) quality model, and in accordance with the Peruvian Technical Health Standard (Norma Técnica de Salud [NTS]). METHODS: A 3-stage methodological study was conducted following COSMIN (Consensus-based Standards for the Selection of Health Measurement Instruments) guidelines. First, a 23-item instrument was developed, grouped into 7 usability dimensions based on ISO 25010. Second, content validity was assessed using a modified Delphi method with 10 experts, using Aiken's V coefficient. Third, structural validity and reliability were evaluated in 115 participants (physicians, nurses, and administrative staff) using confirmatory factor analysis and Cronbach α. Additionally, the module's functional compliance with the NTS was verified. RESULTS: The instrument demonstrated excellent psychometric properties among 115 participants (53 physicians, 48 nurses, and 14 administrative staff); high content validity (Aiken's V=0.95; 95% CI 0.84-0.98), excellent internal consistency (α=.968), and optimal factor fit (Comparative Fit Index=1.000; Tucker-Lewis Index=1.001). Overall usability was perceived as very high by 94 of 115 users (81.7%), and high by 21 of 115 users (18.3%), with no perceptions at lower levels. The module met 28 of 35 requirements (80%) of the Peruvian Technical Health Standard. Administrative staff reported higher satisfaction (13/14, 93% rated it as very high) compared with clinical professionals (81/101, 80% rated it as very high), particularly in the dimensions of error protection and operability. CONCLUSIONS: This study provides the first fully psychometrically validated instrument in Spanish based on ISO 25010, System Usability Scale, and other theoretical foundations for evaluating the usability of information systems. The results highlight the importance of considering both international standards and local regulatory requirements (NTS) in the design of digital health systems. The instrument is applicable in Spanish-speaking contexts and can serve as a reference for future usability evaluations in the region as well as in low- and middle-income countries.
23 May 2026
Read appraisal →Journal of medical Internet research
Human-AI Interaction in Kidney Transplant Decision Support Systems: Qualitative Study of Patient and Support Person Expectations
BACKGROUND: Artificial intelligence (AI) is increasingly applied in medicine, including clinical decision-making. AI-based decision support systems (DSS) can enhance early risk detection and treatment optimization. However, the perspectives of patients and their support persons on AI-assisted DSS in clinical care, particularly regarding shared decision-making (SDM), remain underexplored. OBJECTIVE: This study investigates the expectations, informational needs, and perceptions of patients who underwent kidney transplantation and their support persons regarding AI-assisted DSS and its influence on SDM in posttransplant care. METHODS: In a longitudinal qualitative study, 36 semistructured interviews were conducted with patients who underwent kidney transplantation and their support persons at a German kidney transplant center. Participants were asked about their views on AI's role in follow-up care, its impact on communication, trust, and decision-making, as well as their informational needs regarding AI-assisted DSS. Interviews were transcribed, pseudonymized, and analyzed using framework analysis. RESULTS: Participants recognized AI's potential to support clinicians by identifying risks of transplant loss, rejection, and infection, and by providing data-driven treatment recommendations. However, they emphasized that final decisions should remain with physicians. A majority of participants (n=28, 78%) expressed concern that AI might depersonalize care and diminish physician-patient communication due to a lack of "human touch." Participants demonstrated limited understanding of AI-based DSS functionality and highlighted the need for simple, accessible educational materials (eg, leaflets) explaining AI operations. While most doubted AI could replicate human empathy, some acknowledged that AI might be perceived as more attentive than time-pressured physicians, offering consistent monitoring and support. Participants consistently stressed that AI should augment, not replace, clinical decision-making. CONCLUSIONS: Patients who underwent kidney transplantation and support persons endorse the integration of AI in follow-up care when it enhances clinical decision-making without supplanting the physician's role. Acceptance and trust depend on transparency, accountability, and preserving the "human touch" in care. The development of educational tools to communicate AI functions and limitations is crucial to empower patients and support persons in SDM processes and to ensure AI complements, rather than undermines, patient-centered care.
23 May 2026
Read appraisal →Journal of radiological protection : official journal of the Society for Radiological Protection
The integration of artificial intelligence in radiation medical physics: insights from an international survey with regional variability
Artificial intelligence (AI) is considered to be a leading technology in radiation medical physics, which has the potential for improving efficiency and precision in imaging, radiotherapy, and nuclear medicine. Nonetheless, its application in the clinical setting is hindered by education, regulation, and ethical issues. To evaluate the views of medical physicists on AI adoption, potential issues, perceived advantages, ethical issues, and training needs. A cross-sectional survey in the form of an online questionnaire was distributed internationally to practicing medical physicists, representing specialties in radiation therapy, diagnostic imaging, and nuclear medicine. The instrument was used to capture demographic variables and AI familiarity, perceived benefits and barriers, ethical issues, training preferences, and future expectations. Descriptive statistics, chi-square tests, and binary logistic regression were performed. Open-ended responses were analyzed using thematic analysis. Most respondents were moderately familiar with AI, but nearly half had not incorporated AI into their practice despite being interested. Adoption patterns differed significantly across geographic regions, with higher usage reported in developed countries. AI familiarity was strongly associated with adoption. The most prominent benefits noted were dose calculation accuracy, patient safety, and workflow efficiency. In contrast, the main obstacles were inadequate training, high implementation costs, absence of standardized protocols, and low availability of advanced equipment. Ethical apprehensions focused on accountability and reduced human oversight. Preferred educational strategies included hands-on workshops and on-the-job training. Regional analysis and modeling revealed variability in AI adoption and identified key predictors, including AI familiarity and specialization. Results indicate that AI is considered an important technology in radiation medical physics. However, implementation was influenced by regional and professional differences. These findings highlight the necessity to expand AI education, create standard training programs, improve infrastructure, and provide clearer governance frameworks to support safe and responsible AI integration.
22 May 2026
Read appraisal →Scientific reports
A human-centric fuzzy decision support system for medical diagnosis using fuzzy cognitive maps
Medical decision support requires models that remain interpretable under uncertainty while still adapting to evolving data and expert knowledge. Fuzzy cognitive maps (FCMs) are attractive in this setting because they encode concept-level relations in a transparent graphical form. However, static expert-defined FCMs are often too rigid, whereas purely data-driven updates may weaken semantic consistency and drift away from clinically meaningful relations. This paper proposes a human-centric adaptive framework for fuzzy cognitive map-based medical decision support. The proposed Human-Centric Fuzzy Decision Support System (HCFDSS) combines data-driven weight adjustment with an explicit expert-correction operator, allowing the causal structure of the map to be refined without discarding domain knowledge. The method is formulated through a two-stage procedure: an inner reasoning loop updates concept activations for fixed weights, while an outer learning loop updates the weight matrix using a regularized objective together with bounded expert intervention. Under standard smoothness and boundedness assumptions, we establish sufficient conditions for stability of the inner reasoning dynamics and convergence of the outer weight-update operator. A numerical illustration shows how expert feedback can revise clinically important edges and alter the final diagnostic activation in an interpretable way. In addition, the paper reports a reproducible empirical evaluation involving synthetic medical scenarios and three public medical benchmark datasets from the UCI Machine Learning Repository, together with sensitivity analysis for expert intervention, preprocessing validation, robustness checks under missingness and imbalance, and expanded comparisons with mainstream machine-learning baselines. The proposed HCFDSS is positioned not as a replacement for high-capacity black-box predictors, but as a transparent and editable decision-support framework in which clinicians can inspect, question, and refine the learned relations. The main contribution of the present work is therefore methodological and theoretical, while the empirical evaluation clarifies the predictive-interpretability trade-off of the proposed framework under reproducible benchmark settings.
20 May 2026
Read appraisal →BMJ open
What are the views of cancer care administrators and clinicians in England on the use of a machine learning clinical decision support system (ML-CDSS) to predict patients' risk of hepatic and renal deterioration during chemotherapy? A qualitative study.
OBJECTIVE: To explore administrators' and clinicians' views on the factors that influence their use and adoption of a machine learning clinical decision support system (ML-CDSS) to predict patients' risk of hepatic and renal deterioration during chemotherapy. METHODS AND ANALYSIS: This was a qualitative study that used purposive sampling. 18 participants with administration and clinical backgrounds working in cancer care in England were recruited. Qualitative data were collected by conducting semi-structured interviews and a focus group. Data were analysed thematically using the framework method to identify key themes. RESULTS: Participants acknowledged that monitoring blood chemistry is a core component of chemotherapy as it helps clinicians assess patient fitness and treatment response. The ML-CDSS was perceived as a potentially valuable tool for identifying patients at increased risk of hepatic and renal deterioration, supporting clinical decision-making and enhancing care efficiency. However, several concerns were raised regarding its potential implementation in practice. Participants questioned clinicians' willingness and capacity to integrate the tool into their existing workflows. Participants also believed it was important to demonstrate the ML-CDSS's sensitivity, specificity and validity in accurately predicting patients' risk to build clinicians' trust in the tool, demonstrating evidence of its efficacy and effectiveness in practice. CONCLUSION: Administrators and clinicians recognised the potential benefits of the ML-CDSS to enhance the delivery of chemotherapy by identifying patients at risk for hepatic and renal deterioration. Successful adoption in practice depends on building trust with the tool by being transparent in its development, its effectiveness and impact. Future work should demonstrate the ML-CDSS being used in practice to generate real-world evidence.
20 May 2026
Read appraisal →Philosophical transactions. Series A, Mathematical, physical, and engineering sciences
Revisiting Rogers' Paradox in the context of human-AI interaction
People learn about the world in many ways: from conducting experiments to copying others' behaviour. The choices we make about how to learn can impact the collective understanding of a whole population, were others to learn from us. Alan Rogers developed simulations to study these phenomena-where agents could individually or socially learn amidst a dynamic, uncertain world-and uncovered a surprising result: the availability of cheap social learning yielded no benefit to population fitness over individual learning. Rogers' Paradox spawned decades of work to understand factors that favour social learning and better model human cultural development. But what happens when humans can learn from artificial intelligence (AI) systems that are themselves learning from us? We revisit Rogers' Paradox in the context of human-AI interaction and extend the simulations towards a simplified network of humans and AIs learning together about an uncertain world. We examine the impact of several learning strategies on the equilibrium of a society's 'collective world model' and assess levers available to stakeholders in human-AI interactions to change network dynamics. We then model negative feedback loops that may arise from humans learning socially from AI, and consider other open questions that could be explored in our simulation framework. This article is part of the theme issue 'World models in natural and artificial intelligence'.
17 May 2026
Read appraisal →PloS one
Data-driven differentiation analysis of urban high-tech industries: Research on bibliometrics and large language models.
This study examines inter-city heterogeneity in China's high-tech industries from a regional innovation systems (RIS) perspective, with a particular focus on how variations in knowledge production, technological application, and actor configurations are associated with divergent urban innovation trajectories. We compile more than 39,000 publications from the Web of Science (WOS) and nearly 10,000 patent records from the national patent database for the period 2016-2025, covering four representative cities-Wuhan, Chengdu, Hangzhou, and Tianjin-and four technological domains: artificial intelligence (AI), fiber-optic communication (FOC), intelligent connected vehicles (ICV), and storage chips (SC). The study develops an integrated analytical framework combining bibliometric analysis, co-word network modeling, collaboration network mapping, and large language model (LLM)-assisted semantic interpretation. LLMs are employed primarily in keyword cleaning, terminology standardization, and topic identification, improving the consistency and interpretability of textual metadata. Visualizations generated using VOSviewer highlight pronounced inter-city differences in technological portfolios, research priorities, and collaboration structures. The results suggest distinct urban innovation configurations across the four cities. Wuhan exhibits strong positioning in FOC and SC, reflecting a combined industry-academy orientation. Hangzhou shows high frontier intensity in AI and ICV, consistent with an industry-led and digitally driven innovation profile. Chengdu demonstrates substantial academic output but comparatively weaker evidence of technological translation, while Tianjin, despite a smaller overall scale, displays notable specialization in applied domains such as brain-computer interfaces and smart port technologies. Rather than replacing quantitative analysis, LLM-assisted interpretation supports the identification and contextualization of these patterns by enhancing semantic coherence and reducing noise in large-scale textual data. Overall, the proposed framework provides a reproducible and scalable approach for examining regional technological differentiation and is applicable to comparative studies of urban innovation systems across different regions and industrial contexts.
16 May 2026
Read appraisal →JMIR mental health
Adoption of Digital Mental Health Interventions in National Health Service England, Scotland, and Wales: Freedom of Information Questionnaire Study
BACKGROUND: Digital mental health interventions (DMHIs) have been widely promoted to improve access to mental health care within the UK National Health Service (NHS), particularly following the COVID-19 pandemic. In 2015, a total of 48 technologies were reportedly used in NHS services in England, but over the past decade, substantial changes to regulatory requirements, evidence standards, and procurement processes have reshaped the digital mental health landscape. There is limited clarity regarding which DMHIs are currently being formally procured and funded by NHS mental health services across the United Kingdom. OBJECTIVE: This study aimed to identify and describe the DMHIs currently procured, contracted, or paid for by NHS mental health service providers in England, Scotland, and Wales for adult common mental health problems and to compare current procurement practices with findings reported in 2015. METHODS: Freedom of Information requests were submitted to all NHS mental health trusts in England and all health boards in Scotland and Wales. Responses were collated and screened to provide an updated and extended record of which technologies are reportedly procured or paid for by services. RESULTS: In total, 19 different DMHIs were identified as being procured across mental health service providers for adult common mental health problems at the time of data collection. This demonstrates a substantial reduction in the number of technologies being adopted into practice compared to the 48 reported in England in 2015. The findings reveal several key insights, including that only 2 technologies have remained in use for a decade, and they shed light on the types of technologies being selected and the variations in procurement practices among the 3 national health services. CONCLUSIONS: Despite the expansion of the digital mental health marketplace, the number of DMHIs formally procured by NHS mental health services has markedly decreased over the past decade. This consolidation may reflect increased selectivity and the adoption of higher-quality products, driven by strengthened regulatory oversight, evidence standards, and national guidance. Although these developments may enhance safety and quality assurance, they also raise important questions about innovation, market sustainability, and equitable access to digital mental health care. Ongoing monitoring of procurement practices is needed to inform policy, service design, and the future development of DMHIs.
16 May 2026
Read appraisal →JMIR research protocols
Evaluating the Methodological Quality of Artificial Intelligence-Assisted Systematic Reviews: Protocol for a Mixed Methods Meta-Research Study
BACKGROUND: Artificial intelligence (AI), including large language models (LLMs), is increasingly integrated into systematic review (SR) workflows. AI tools may accelerate searching, screening, data extraction, and reporting, but their effects on methodological quality, reporting completeness, transparency, and reproducibility remain uncertain. Existing evaluations largely examine isolated tasks, and inconsistent disclosure of AI use limits reproducibility and oversight. OBJECTIVE: This 4-phase mixed methods meta-research study will (1) compare the methodological quality of AI-assisted versus traditional SRs; (2) refine, finalize, and apply a preliminary AI Transparency and Disclosure Index (AITDI); (3) evaluate reproducibility by comparing outputs across repeated runs of the same AI model, across different AI models, and between AI models and human reviewers at multiple SR stages; and (4) explore knowledge user perspectives on rigor, transparency, and trust in AI-assisted SRs. METHODS: We will conduct a matched cohort analysis of SRs published from 2023 to 2025 in biomedical journals. Each AI-assisted SR will be matched 1:2 with traditional SRs by publication year, clinical domain, review type, and meta-analysis status. Two independent reviewers will apply A Measurement Tool to Assess Systematic Reviews, version 2 (AMSTAR 2; methodological quality), PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 (reporting completeness), and, when applicable, Risk of Bias in SRs (ROBIS; risk-of-bias rigor). A preliminary AITDI will be refined and then applied to all AI-assisted SRs. Reproducibility will be assessed using SR-derived task sets to compare outputs across repeated runs of the same model, across different models, and between AI and human reviewers at key SR stages. Semistructured interviews with authors, editors, clinicians, policymakers, and patient partners will be analyzed using reflexive thematic analysis. RESULTS: As of December 2025, the study has been preregistered on the Open Science Framework (OSF; DOI: 10.17605/OSF.IO/Q5JRW), the search strategy has been finalized, and title/abstract screening has begun. Data extraction is planned for March-May 2026, followed by AITDI refinement and reproducibility testing from May 2026 to October 2026. Qualitative interviews are anticipated from October 2026 to February 2027, with final analyses by April 2027 and dissemination planned for mid-2027. CONCLUSIONS: This study will provide one of the first empirical comparisons of methodological quality, transparency, and reproducibility of AI-assisted versus traditional SRs in the LLM era. Findings will inform expectations for responsible AI integration and support refinement of reporting and methodological best practices, including future development of AI-specific reporting and appraisal extensions (eg, PRISMA-LLM [Preferred Reporting Items for Systematic Reviews and Meta-Analyses-large language model] and AMSTAR-LLM [A Measurement Tool to Assess Systematic Reviews-large language model]).
16 May 2026
Read appraisal →PloS one
Classification of soft decision-making methods via fuzzy parameterized fuzzy soft matrices and their performance-based statistical analysis in machine learning
This study provides a comprehensive evaluation and classification of 35 soft decision-making (SDM) algorithms based on fuzzy parameterized fuzzy soft matrices (fpfs-matrices). Although fpfs-matrices offer a strong mathematical framework for modeling uncertainty, there has been a lack of large-scale comparisons of their derivative SDM methods in machine learning. To address this, we used the Comparison Matrix-Based Fuzzy Parameterized Fuzzy Soft Classifier (FPFS-CMC) to benchmark these 35 algorithms across ten diverse datasets from the UCI Machine Learning Repository. The methods were thoroughly assessed utilizing metrics such as accuracy, precision, recall (sensitivity), specificity, and F1-score, and statistical significance was confirmed using the Friedman and Nemenyi tests. Our results show that SDM methods via fpfs-matrices perform competitively in classification tasks involving uncertainty. Notably, the best algorithms according to F1-scores were A19 (Rank 1), YHX14 (Rank 2), and a three-way tie for Rank 3 among VMH16, AKO18o, and A19/2. By identifying the most effective algorithms and offering a structured decision-support framework, this research provides both a theoretical reference and practical guidance for practitioners selecting SDM methods for complex machine learning challenges.
15 May 2026
Read appraisal →Journal of medical Internet research
AI-Based Automation for Medication Reconciliation: Scoping Review
BACKGROUND: Medication reconciliation (MedRec) has the potential to improve patient safety by enhancing the continuity of medication information across settings. MedRec involves 3 core tasks: the creation of a best possible medication history, the identification of medication discrepancies among medication lists, and the resolution of medication discrepancies. While artificial intelligence (AI) has the potential to improve MedRec, existing reviews have not identified the ways in which researchers have used AI to facilitate MedRec tasks and their constituent subtasks or the level of automation achieved. OBJECTIVE: This scoping review aimed to map how previous research has applied AI to MedRec tasks and subtasks and assess the extent of automation achieved. METHODS: We searched MEDLINE, Embase, Web of Science, IEEE Xplore, and Compendex in June 2024 for studies that used AI to support a MedRec task or subtask, excluding entirely rule-based tools or studies focused on other aspects of medication management. After screening 2345 unique records, we conducted backward citation searching of studies included at the full-text stage, identifying an additional 795 unique records. We used a 4-stage model of human information processing as a structural lens to guide our considerations of automation, mapping the core tasks of MedRec onto this model. RESULTS: A total of 94 studies met the inclusion criteria. All studies addressed subtasks related to the creation of a best possible medication history. Only 2.1% (n=2) of the studies also addressed the identification of discrepancies. Thus, the highest stage of automated information processing achieved was information analysis, although most studies (92/94, 97.9%) only automated information acquisition steps. Most studies (67/94, 71.3%) used free-text clinical notes from the electronic health record, although a significant proportion (21/94, 22.3%) used images of pills or images of other medication-related items. Studies using text-based data used a variety of machine learning methods (eg, recurrent neural networks, conditional random fields, support vector machines, and transformers), whereas those that leveraged images typically used convolutional neural networks. Most studies (61/94, 64.9%) used publicly available data from benchmarking datasets (eg, n2c2 2022) and were strictly model development studies, with only 1.1% (n=1) being usability studies. CONCLUSIONS: This is the first review to consider the role of AI in the automation of MedRec tasks, offering a basis for prioritizing future development efforts. Current applications of AI to automate MedRec tasks are preliminary, with most work focusing on the extraction of medication information and limited to proof-of-concept model development. Future work should consider addressing infrastructural barriers to the AI-based automation of MedRec tasks (eg, data incompleteness in sources of medication information) and exploring approaches to automate discrepancy resolution. Beyond developing models, there is also a need to implement them in tools and evaluate them in real-world contexts.
13 May 2026
Read appraisal →Journal of medical systems
Clinical Plausibility in Large Language Model Robustness Testing for Medicine: A Scoping Review
Large language models (LLMs) show promise in medical applications, yet their translation into clinical practice requires rigorous validation. Current robustness testing often employs adversarial approaches borrowed from AI safety, raising questions about their alignment with authentic clinical scenarios. To systematically map methodologies used for robustness testing of LLMs in medical contexts and assess their clinical plausibility. A scoping review was conducted following PRISMA-ScR guidelines, searching PubMed, Embase, Web of Science, IEEE Xplore, ACM Digital Library, arXiv, and MedRxiv from January 2023 to September 2025. Two independent physician reviewers screened 5,331 articles, extracting data on testing methodologies, medical domains, expert involvement, and clinical plausibility. Thirty-three studies met inclusion criteria, predominantly from 2025 (82%). The most common robustness testing approaches were misleading prompts (49%) and adversarial prompts (39%). Only 33% of studies designed tests clearly mimicking plausible clinical scenarios. While 58% reported expert involvement, the depth of integration varied considerably. Studies predominantly addressed mixed medical domains (73%) rather than specialized fields. The emerging literature suggests that LLM robustness testing in medicine often emphasizes technical vulnerability detection, with fewer studies examining clinically plausible scenarios of routine use. Future frameworks should complement adversarial testing with clinically grounded, longitudinal, and specialty focused evaluations to support deployment-relevant inference.
13 May 2026
Read appraisal →Integrative psychological & behavioral science
The Human Mind in Dialogue With its Replica AI - a Historical Perspective
Although artificial intelligence has exploded for the last couple of years, the idea of comparing the human mind with a computer is not new. One may say that Alan Turing started this movement, and this idea got off to a flying start in psychology through the symposium arranged by the "Special Interest Group in Information Theory" at Massachusetts Institute of Technology Sept. 10-11, 1956. This symposium was sponsored by IBM, and Noam Chomsky presented for the first time his ideas about a transformational generative grammar. There were several contributors to the cognitive revolution, but Chomsky's presentation of a linguistic system with a level of precision comparable with mathematics was symptomatic for how a cognitive model for the human mind aimed to be conceptualized. This has brought the question about artificial intelligence to the core of psychology, namely how it can be understood as a model for the human mind. Psychology and its understanding of the human mind have changed along the last 500 years of the Western psychology. After a short presentation of some of the premises for artificial intelligence of today and some aspects of the idea behind the computational mind, these perspectives are compared with the very early use of the term 'psychology' in the early 16th century. The role of classical logic as a model for the human mind was at stake back then and one of the main factors that paved the way for psychology was that Aristotelian logic was criticized and questioned as a model for the human mind. On the other hand, the importance of classical logic has never been completely outdated Consequently, for the last 500 years Western psychology can in many ways be characterized as an incessant battle about the role of logic in the understanding of the human mind. However, the conclusion is that AI of today presupposes an understanding of the human mind that has been outdated for more than 500 years.
13 May 2026
Read appraisal →BMC medical research methodology
When to stop reviewing: validation of stop criteria in ASReview.
BACKGROUND: Systematic reviews are essential for evidence-based research, but they often require a great deal of time and effort. Although title and abstract (T&A) screening is just one part of the review process, it can be very time-consuming when search strategies retrieve large numbers of records. Given the exponential growth of scientific publications in recent decades, tools such as ASReview, which use machine learning (ML) for active learning-based screening, aim to reduce the workload. However, since ASReview helps users prioritise potentially relevant studies rather than supporting them in screening all records, a key question arises: at what point can the screening process safely stop without risking the omission of relevant studies when the entire dataset is not being reviewed? METHODS: This simulation study tested three proposed stop criteria for terminating screening in ASReview without loss of relevant data: (1) stopping after a calculated number of relevant studies based on an initial sample; (2) stopping after a fixed number of consecutively studies deemed irrelevant; (3) stopping after a predefined percentage of the dataset has been screened. A total of 35,000 automated title and abstract screenings were conducted using five datasets from the SYNERGY repository. Key outcomes included the percentage of studies screened until the last relevant study was found and the number of relevant studies missed under each stop criterion. RESULTS: The proportion of the dataset that needed to be screened to identify all relevant studies (as pre-classified in the SYNERGY dataset) varied greatly across datasets, ranging from 2.9% to 76.9% on average. None of the tested stop criteria could consistently identify all relevant studies across all datasets. Stop criterion 1 was reliable in only 2% of simulations. Stop criterion 2 showed high variability, with thresholds ranging from 2% to 61%, depending on the dataset. Stop criterion 3 failed to define a universal percentage applicable across datasets. CONCLUSIONS: ASReview can reduce screening workload by prioritizing potentially relevant studies through ML-based ranking, thereby allowing researchers to identify relevant studies earlier in the screening process. However, no stop criterion reliably ensures that all relevant studies are identified. Early stopping may result in missed studies, depending on dataset characteristics. Current stop criteria should be applied cautiously and potentially combined with quality assurance measures. Further research is needed to develop more robust and generalizable stopping rules. TRIAL REGISTRATION: Not applicable - this is a simulation study, not a registered systematic review.
13 May 2026
Read appraisal →Scandinavian journal of primary health care
Identifying factors influencing acceptance of artificial intelligence among general practitioners in Danish general practice: a cross-sectional web-based survey study
BACKGROUND: Despite rapid advances in artificial intelligence (AI), its adoption in Danish general practice remains limited and decentralized, relying on individual general practitioners' (GPs') decisions. This study aimed to evaluate Danish GPs' acceptance of AI and the importance of factors derived from established technology-acceptance models. METHODS: A cross-sectional web survey was conducted from mid-September 2024 to March 2025, using a culturally adapted version of a previously validated tool. The survey included 42 items: nine background items and 33 items covering 11 factors, including Medical and Non-Medical Performance Expectancy, Effort Expectancy, Social Influence (Medical and Patient), Facilitating Conditions, Perceived Trust, Anxiety, Professional Identity, Innovativeness and Behavioral Intention. Descriptive statistics, subgroup comparisons and Cronbach's alpha were provided. RESULTS: A total of 109 GPs and residents responded, with 84.4% (n = 92) completing the survey. Attitudes toward AI were generally neutral to positive, with mean scores above neutral in seven of the 11 factors. The most positive factors measured were Medical and Non-Medical Performance Expectancy. Scores related to Behavioral Intention were also high. Lower scores appeared in Perceived Trust, Facilitating Conditions and Anxiety. Older GPs (60 years and above) reported lower scores on Non-Medical Performance Expectancy and Social Influence (Medical); GPs in smaller cities showed more positive attitudes across several factors. Cronbach's alpha indicated good internal consistency on all but two scales. CONCLUSION: In summary, Danish GPs demonstrate a strong intention to adopt AI. However, issues persist as significant barriers regarding Anxiety, Perceived Trust and Facilitating Conditions. Implementation strategies considering context and clinician demographics are recommended.
11 May 2026
Read appraisal →PloS one
Technological impact, generality, and complementarity of artificial intelligence patents: Evidence from Samsung Electronics.
This study examines the extent to which artificial intelligence (AI) technologies are more pervasive and technologically complementary than non-AI technologies. Drawing on the literature on general purpose technologies (GPTs), we hypothesize that AI technologies exhibit greater pervasiveness-reflected in higher technological impact and broader generality-and stronger technological complementarity than non-AI patents. Using a firm-level analysis of patents filed by Samsung Electronics between 1982 and 2018, we find that AI patents demonstrate significantly higher impact, wider generality, and stronger systematic complementarity compared to matched non-AI patents. These results are robust across alternative matching specifications and multiple operationalizations of the key variables. Our findings offer firm-level empirical supporting AI's GPT-like characteristics and shed light on how AI-driven innovations may generate sustained incentives for technological advancement, even amid processes of creative destruction. The paper concludes by discussing key theoretical, managerial, and policy implications.
11 May 2026
Read appraisal →JMIR formative research
Older Adults' Experiences Navigating Setup of Digital Health Technology: Implementation Report
BACKGROUND: Digital health and connected technologies may support better health outcomes among older adults, including those with multiple chronic conditions or low engagement in health behaviors. However, initial experiences with technology, including during unboxing, setup, and first use, can influence emotional reactions and perceptions and can ultimately determine sustained, meaningful use. Older adults with low technology experience or poor health may be particularly vulnerable to frustration, stress, or abandonment of devices when early interactions are negative. OBJECTIVE: The purpose of this implementation study was to closely observe the initial engagements with a telehealth treatment app and connected blood pressure monitor (BPM) among a group of older adults with low prior technology use and reported low health behavior engagement. The goal was to identify setup "pain points" that may influence initial impressions and intention to use the technology over time. METHODS: A total of 24 older adults (aged ≥65 years) were recruited for a 4-week trial of a telehealth app. Participants were provided with a box containing a tablet preloaded with the app, paper instructions, and a BPM and cuff. Researchers first conducted in-home ethnographic interviews with participants to observe the unboxing and setup process, documenting experiences with reading instructions, using the BPM, and engaging with customer support. Weekly check-in calls and a final exit interview captured ongoing experiences and likelihood of continued use. Interview recordings were transcribed and independently coded, guided by the unified theory of acceptance and use of technology. RESULTS: Most of the sample were White (20/24, 83%) and female (14/24, 58%). Negative experiences with the app's customer support were the top challenge for participants, with representatives providing confusing steps or conflicting terminology. Other common challenges were understanding instructions, connecting to Bluetooth, and correctly using the BPM. While 67% (16/24) of the participants indicated that they were likely or very likely to continue to use the app after the study ended at the end of week 1, this number dropped to 54% (13/24) by the end of the 4 weeks. Participants who reported lower technology self-efficacy at the beginning of the study also experienced frustration, anxiety, and embarrassment as friction with the setup process continued. CONCLUSIONS: First impressions of digital health apps play a critical role in influencing older adults' emotions and perceptions regarding the technology and may impact the likelihood of longer-term engagement. Those with lower technology self-efficacy are particularly susceptible to experiencing negative emotions such as frustration, stress, or shame. Mobile health apps and interventions targeting older adults should incorporate simplified instructions with clear, consistent terminology and well-trained customer support staff to improve the onboarding experience.
11 May 2026
Read appraisal →Therapeutic innovation & regulatory science
Applications and Challenges of Artificial Intelligence and Big Data in Drug Regulation
Artificial intelligence (AI) and big data are increasingly applied in drug regulation and have demonstrated significant potential worldwide. The U.S. Food and Drug Administration (FDA) has developed a relatively comprehensive approach through strategic frameworks, regulatory guidelines, and pilot programs. The European Medicines Agency (EMA) has promoted AI adoption via the Big Data Task Force, DARWIN EU®, and a multi-annual work plan, while Japan, Canada, and other countries have also advanced relevant initiatives and strengthened international cooperation. In China, smart regulation has been incorporated into the "14th Five-Year Plan" and subsequent strategies, with progress in establishing national regulatory data platforms, pharmaceutical traceability systems, and pilot AI applications. Nevertheless, AI in drug regulation remains at an exploratory stage, facing challenges such as limited model reliability and interpretability, insufficient data standards and interoperability, regulatory gaps, and ethical as well as public trust concerns. Future progress will depend on strengthening regulatory standards, enhancing data governance, improving regulatory capacity, and deepening international collaboration to achieve more scientific, intelligent, and efficient drug regulation.
9 May 2026
Read appraisal →Journal of biomedical informatics
Model utility and explainability in federated learning - A case study in healthcare using fundus oculi datasets
OBJECTIVE: Introduce a case study for Federated Learning (FL) in healthcare, addressing challenges posed by patient privacy and limited large-scale datasets. Our goal is to assess the features learned by FL methods in a simulated, diverse setting that emphasizes realistic data heterogeneity, and to analyze the learned representations for their medical relevance using both local and global explainability techniques. METHODS: Six fundus oculi datasets were combined to simulate a diverse federated learning environment, representing heterogeneous data conditions. We evaluated three established FL methods against centrally trained models, assessing both predictive performance and the learned representations. Specifically, explainability techniques were employed to examine the features learned by the models, and local explanations were evaluated against attention maps annotated by ophthalmologists. Robustness against common biases in fundus datasets was also assessed. RESULTS: Our study found improvements in model utility (up to 9.97%) with FL methods compared to isolated training. Analysis of learned representations revealed that federated models predominantly learn the vertical cup-to-disc ratio, a crucial feature for glaucoma diagnosis, and demonstrated robustness against common biases. High agreement was observed between local explanations and ophthalmologist-annotated attention maps. CONCLUSION: This study demonstrates the benefits of FL systems in a healthcare scenario, providing a case study for evaluating federated systems beyond idealized benchmarks. Our findings highlight the potential of FL to not only improve model utility in privacy-sensitive medical domains but also to learn medically relevant features instead of spurious correlations.
9 May 2026
Read appraisal →Research synthesis methods
Transforming evidence synthesis: A systematic review of the evolution of automated meta-analysis in the age of AI
Exponential growth in scientific literature has heightened the demand for efficient evidence-based synthesis, driving the rise of the field of automated meta-analysis (AMA) powered by natural language processing and machine learning. This PRISMA systematic review introduces a structured framework for assessing the current state of AMA, based on screening 13,216 papers (2006-2024) and analyzing 61 studies across diverse domains. Findings reveal a predominant focus on automating data processing (52.5%), such as extraction and statistical modeling, while only 16.4% address advanced synthesis stages. Just one study (approximately 2%) explored preliminary full-process automation, highlighting a critical gap that limits AMA's capacity for comprehensive synthesis. Despite recent breakthroughs in large language models and advanced AI, their integration into statistical modeling and higher-order synthesis, such as heterogeneity assessment and bias evaluation, remains underdeveloped. This has constrained AMA's potential for fully autonomous meta-analysis (MA). From our dataset spanning medical (67.2%) and non-medical (32.8%) applications, we found that AMA has exhibited distinct implementation patterns and varying degrees of effectiveness in actually improving efficiency, scalability, and reproducibility. While automation has enhanced specific meta-analytic tasks, achieving seamless, end-to-end automation remains an open challenge. As AI systems advance in reasoning and contextual understanding, addressing these gaps is now imperative. Future efforts must focus on bridging automation across all MA stages, refining interpretability, and ensuring methodological robustness to fully realize AMA's potential for scalable, domain-agnostic synthesis.
6 May 2026
Read appraisal →Drug development research
Artificial Intelligence in Drug Discovery and Development: Transforming Pharmaceutical Innovation
Drug discovery remains a lengthy, costly, and high-risk endeavor, often requiring over a decade from target identification to clinical translation. Artificial intelligence (AI) is reshaping this paradigm by enabling more efficient and accurate decision-making across the discovery and development pipeline. Advances in machine learning, deep learning, and natural language processing now support target identification, hit finding, lead optimization, and drug repurposing with unprecedented speed and precision. AI-driven insilico platforms further enhance early-stage predictability by forecasting toxicity, pharmacokinetics, and developability, thereby reducing late-stage attrition. This review critically examines the evolving role of AI in modern drug discovery and its expanding impact on pharmaceutical formulation development and personalized medicine. Collaborative models between AI developers and the pharmaceutical industries, essential for accelerating translational outcomes, are also highlighted. Finally, key challenges, including algorithmic transparency, data quality, interoperability, and regulatory acceptance, are discussed, along with future directions for harnessing AI's full potential in pharmaceutics.
6 May 2026
Read appraisal →Radiology. Imaging cancer
ONCO-RADS-guided Large Language Models for Extraction and Classification of Incidental Findings on Whole-Body Imaging Reports
Purpose To evaluate large language model (LLM)-based strategy performance for extraction and classification of incidental findings from whole-body (WB) imaging reports, particularly strategies incorporating Oncologically Relevant Findings Reporting and Data System (ONCO-RADS). Materials and Methods In this retrospective bicenter study, authors included all WB MRI reports from January 2016 to December 2023 at a referral center (internal dataset). Two observers extracted all incidental findings, and patient records were used to confirm final diagnoses. First, authors evaluated ONCO-RADS performance and the reproducibility of its incidental finding classifications by six radiologists. Then, authors evaluated the accuracy of three LLM-based strategies: (a) a fine-tuned DeBERTa/medical named entity recognition (NER) model; (b) zero-shot LLMs (ChatGPT-o1 [OpenAI], Gemini-2.5-Pro [Google]); and (c) reference-guided prompting of these LLMs using ONCO-RADS. Authors then expanded these strategies to an external dataset of 605 reports with multiple imaging techniques (405 WB MRI; 100 fluorodeoxyglucose PET/CT; and 100 chest-abdomen-pelvis CT acquisitions) from January 2022 to January 2025. Results The internal dataset included 823 patients (mean age, 63.7 years ± 11.7 [SD]; 457 male patients) with 1488 WB MRI reports. The average interobserver reproducibility of ONCO-RADS incidental finding classifications was excellent (Cohen κ, 0.87). The per-report accuracies of ONCO-RADS-guided LLMs (95.6% [151 of 158] and 86.7% [137 of 158] for ChatGPT-o1 and Gemini-2.5-Pro, respectively) were higher than those of the medical NER (69.0% [109 of 158]) and zero-shot LLMs (57.0% [90 of 158] and 70.9% [112 of 158] for ChatGPT-o1 and Gemini-2.5-Pro, respectively) (P < .001). In the external test set (mean age, 60.6 years ± 12.9; 330 male patients), the per-report accuracies of ONCO-RADS-guided ChatGPT-o1 (83.5% [505 of 605]) and Gemini-2.5-Pro (82.0% [496 of 605]) were higher than those of the models without ONCO-RADS prompting (63.1% [382 of 605] and 61.2% [370 of 605], respectively) and the medical NER (55.7% [337 of 605]) (P < .001). Conclusion Reference-guided prompting of the LLMs ChatGPT-o1 and Gemini-2.5-Pro with ONCO-RADS improved their performance in extracting and classifying incidental findings on WB imaging reports compared with zero-shot prompting and medical NER. Keywords: Large Language Models, Incidental Findings, Whole-Body MRI Supplemental material is available for this article. © RSNA, 2026.
6 May 2026
Read appraisal →Scientific reports
Intelligent virtual agents in psychotherapy: a safety evaluation across high-risk mental health scenarios
The growing burden of mental illness and limited access to evidence-based psychotherapy have increased interest in artificial intelligence (AI)-driven conversational agents as potential supports for mental health care. In this exploratory pilot study, we examined the safety and feasibility of an intelligent virtual agent (IVA) designed to simulate psychotherapeutic interactions, with a focus on high-risk situations involving suicidality and substance use. Two licensed psychotherapists engaged in scripted interactions with the IVA across 12 predefined scenarios addressing suicidality and substance abuse. The IVA was powered by GPT-4omni and embedded in a Unity-based avatar. After each interaction, testers evaluated acceptance, usability, and human-robot interaction. Two independent psychotherapists rated the IVA's responses using a structured scale assessing guideline adherence, risk recognition, help provision, de-escalation, and empathy. No real patients were involved; all interactions were simulated for safety testing purposes. The IVA showed preliminary indications of good usability and generally empathic responses. However, problematic responses occurred in 29% of conversations, with 12.5% rated as highly critical. Responses rated as "critical" or "highly critical" referred to outputs that failed to provide adequate support, showed insufficient risk recognition, or included ethically problematic suggestions. Key concerns included inadequate recognition of risk, normalization of substance use, and insufficient referral to crisis resources, particularly in scenarios involving underage alcohol access and suicide-related inquiries. In this small, expert-based pilot safety evaluation, the findings suggest that although AI-based agents may improve access to mental health support, rigorous safety evaluation, clinical oversight, and robust safeguards are essential prior to clinical deployment. No clinical conclusions can be drawn from this simulated study.
1 May 2026
Read appraisal →PloS one
Research on the development of an automated system for psychology questionnaire generation based on large language models
This study reimagined the psychology questionnaire development process using large language model ((LLM) technology, aiming to overcome the protracted preparation cycles and significant human bias inherent in traditional scale development. We developed a specialized fine-tuning scheme for a corpus of 169 professional psychological questionnaires. By integrating instruction fine-tuning with human feedback reinforcement, we significantly enhanced the adaptability of the Qwen-2.5 and GLM-4 models for demanding professional psychological assessment tasks. The optimized models demonstrated remarkable gains across key dimensions: text generation quality (BLEU-4 increased by 0.05, ROUGE-L by 0.057), scientific rigor (logical consistency improved by 28.6%), and cultural adaptability (achieving over 85% accuracy in cross-regional expression conversion). This research solidly supports the feasibility of leveraging LLM technology to drive research paradigm transformation in psychology, offering crucial methodological support for developing efficient, intelligent psychological measurement tools.
27 Apr 2026
Read appraisal →Journal of medical Internet research
Exploring Attitudes Toward AI-Based Contactless Sensors in Health Among Five Stakeholder Groups: Qualitative Study
BACKGROUND: The rapid rise of artificial intelligence-based contactless sensors (AI-CS) is expected to significantly transform how patients are measured, monitored, and understood through a versatile, noninvasive approach to data collection and health assessment. However, there is a lack of empirical research specifically focusing on AI-CS in health. Moreover, existing studies tend to focus on medical or patient perspectives, while neglecting other stakeholders such as researchers, political actors, or the general public. OBJECTIVE: The study aims to provide an in-depth empirical ethical analysis and, through a multistakeholder approach, a uniquely comprehensive overview by addressing the research question: what are the attitudes of different stakeholders (patients, health care professionals, researchers, political stakeholders, and the general public) toward AI-CS and their applications in health? METHODS: We conducted a cross-sectional study with 104 participants using a semistructured interview guide. Interviews were analyzed using qualitative content analysis with ATLAS.ti software (ATLAS.ti Scientific Software Development GmbH), following a 3-component model of feelings, thoughts, and behavioral aspects. RESULTS: The results of the study provide an in-depth analysis of attitudes toward AI-CS in health among different stakeholders. Overall, the results show a high level of openness to AI-CS in health across all stakeholder groups. In terms of feelings and their correlation with behavioral aspects, 2 key trends emerged: first, greater experience and knowledge correlated with a reduced tendency to react emotionally. Second, participants with positive experiences with technologies were generally more open and positive toward contactless sensors. The combined findings on thoughts and behavioral aspects highlighted 3 key tensions-around contact(lessness) and the importance and ambivalence of touch, between protection and surveillance (particularly regarding path- and context-dependency) and between the benefits and challenges of unobtrusiveness (especially in relation to control and governance implications). In addition, the analysis revealed the need for information and consent about AI-CS and clarified possible technical implementations and fields of application. CONCLUSIONS: This study provides a comprehensive and empirically grounded ethical analysis of stakeholder attitudes toward AI-CS in health. The findings offer valuable guidance for the responsible development, implementation, and governance of AI-CS in health care contexts.
27 Apr 2026
Read appraisal →Biomedizinische Technik. Biomedical engineering
AI in healthcare: mapping Romania's transition from research to intellectual property & industry solutions
OBJECTIVES: Romania is aligning its healthcare AI ecosystem with the European Union's AI Act (2024/1689), through its own National AI Strategy. Despite a growing volume of research in medical AI, significant gaps remain in translating this output into patents and domestically developed high-risk medical devices. This study assesses Romania's research productivity, patenting activity, and commercial solutions in medical AI. METHODS: We conducted an analysis combining bibliometric data, patent information, and medical device databases, and analyzed 619 Romanian-authored articles on medical AI, 272 patent records, and identified domestic AI-powered software providers. RESULTS: Romania's publication output in medical AI has surged post-2018, primarily driven by academic institutions in Bucharest and Cluj. Patents are predominantly filed by multinationals, indicating a potential disconnect between research and industrial output. Six AI medical software solutions were identified, most under the low-risk classifications. CONCLUSIONS: While Romania is expanding its medical AI research, it faces barriers in converting academic output into innovation. The dominance of foreign corporations in patent filings, reliance on international funding for high-impact research, and the scarcity of domestically developed high-risk medical AI solutions highlight important gaps. Addressing these disparities is essential for national alignment in the AI medical innovation space.
26 Apr 2026
Read appraisal →The Nurse practitioner
Artificial intelligence for early detection of aortic stenosis: The emerging role of NPs.
Aortic stenosis is associated with significant morbidity and mortality if left untreated. Nurse practitioners in primary care are uniquely positioned to identify early signs of this condition and are often the frontline provider for patients. Unfortunately, barriers to screening, diagnosis, referral, and treatment remain significant. Artificial intelligence technologies are emerging to address these barriers. This article highlights the potential impact of artificial intelligence in transforming aortic stenosis management and explores its integration into nurse practitioner practices to improve outcomes.
25 Apr 2026
Read appraisal →Journal of medical economics
Value frameworks for digital health technologies: a comparative analysis
AIMS: This study aimed to review and compare existing value frameworks for digital health technologies (DHTs) and identify common domains to guide their evaluation. MATERIALS AND METHODS: A comprehensive literature search was conducted across PubMed, Embase, and Google Scholar, along with citation and website searches, was conducted in August 2025. English-language publications that provided guidance on evaluating value of DHTs were included. A narrative review approach was employed to synthesize findings and identify key value domains and indicators. RESULTS: From an initial pool of 1,132 articles, four met the inclusion criteria and formed the basis of the review, supplemented by additional frameworks from relevant organizations. Six core value domains were identified: technical and security aspects, data rights and governance, clinical characteristics, economic characteristics, health inequalities, and user preferences. Five value frameworks from the United Kingdom (UK) National Health Service, Digital Therapeutic Alliance, World Bank, Institute for Clinical and Economic Review-Peterson Health Technology Institute, and UK National Institute for Health and Care Excellence were compared across these domains. A heat map analysis demonstrated variation in the inclusion of indicators, reflecting differing stakeholder priorities and perspectives. LIMITATIONS: Limitations include potential bias, subjectivity, and limited scope inherent in a narrative review, although elements of PRISMA were applied to reduce risks. Non-English frameworks and those specific to certain technologies were excluded. Additionally, frameworks for both DHTs and digital therapeutics were considered together. CONCLUSION: Six core value domains are consistently emphasized across value frameworks. However, variations across value frameworks highlights the influence of stakeholder needs and priorities. Further research is needed to refine these frameworks and support consistent and informed decision-making in DHTs adoption and implementation. Digital health technologies (DHTs) are increasingly used in healthcare, but their value is assessed in variable ways. We reviewed published studies to identify common elements. Six key areas emerged: technical and security aspects, data rights and governance, clinical and economic characteristics, health inequalities, and user preferences. While some areas are consistently included, others vary across frameworks. Our findings suggest that more consistent approaches could support clear decisions about adopting digital health tools.
25 Apr 2026
Read appraisal →BMJ health & care informatics
Barriers associated with the implementation, adoption, scale-up and sustainability of mHealth in Sub-Saharan Africa: a systematic review guided by the NASSS Framework
OBJECTIVES: The objectives of this study were to highlight and evaluate the factors affecting the non-adoption, abandonment of and challenges to scale-up, spread and sustainability of mHealth in Sub-Saharan Africa (SSA) and to assess the appropriateness of the Non-adoption, Abandonment, Scale-up, Spread, and Sustainability (NASSS) Framework in the SSA context. METHODS: A systematic search of published research was conducted using PubMed and Embase research databases from 2012 to 2025. Peer-reviewed studies from SSA countries, evaluating mHealth interventions, were included in this study. Papers were analysed to identify barriers according to the seven domains of the NASSS Framework, which are (1) condition, (2) technology, (3) value proposition, (4) adopters, (5) organisation(s), (6) wider system, (7) embedding and adaptation over time RESULTS: This study identified 409 barriers across 91 papers. 309 barriers were covered by the seven NASSS Framework domains, while 100 were unaddressed by the framework. The main barriers were associated with domain 2, the technical aspects of the technology (27%). Other prominent barriers related to the wider system (18%), the adopter (11%), the organisation (9%) or condition (8%), with the value proposition (1%) and embedding and adaptation over time (1%), least hindering implementation. Challenges which were unaddressed by the framework (24%) included socioeconomic and infrastructural issues, distinct to lower and middle-income countries (LMIC). DISCUSSION: This paper highlights the importance of evaluating rudimentary aspects of mHealth and preparing infrastructure before addressing more complex issues during mHealth implementation in SSA. CONCLUSION: Although the NASSS Framework is applicable to the SSA region, it would benefit from some further development to improve its applicability in LMICs.
25 Apr 2026
Read appraisal →PloS one
Understanding older adults' perception, acceptance, and adoption of smart home technologies
Smart home technologies have increased in popularity and affordability in recent years, however, there is limited research on their adoption specifically among older adults. This study aims to uncover incentives and barriers to the adoption of smart home components, given the impact of COVID-19 on older adults' quality of life (QoL). For this purpose, online audio-recorded interviews were conducted with 21 participants aged between 65 and 90 years from the UK, Malta, and Pakistan. Participants were shown various smart home technology images and videos and asked for their perception, existing knowledge, and current use. The corresponding audio recordings were transcribed and thematically analysed. Findings revealed that older adults experienced a decline in QoL during COVID-19, accompanied by an increased reliance on digital technologies. While participants recognised the potential of smart home devices to enhance independence and well-being, their knowledge of available options was limited. Key barriers included affordability, privacy and trust concerns, compatibility and integration issues, and environmental and social considerations. Cross-cultural differences emerged, with European participants reporting higher familiarity than those in Pakistan, particularly rural areas where availability was limited. These results highlight the need for designers and policymakers to improve affordability, usability, and cultural adaptability in order to support older adults' acceptance and meaningful adoption of smart home technologies.
23 Apr 2026
Read appraisal →PloS one
An explainable multi-head attention network for healthcare IoT threat detection based on the MedDefender-MHAN framework
The rapid proliferation of Internet of Medical Things (IoMT) devices in healthcare environments has created critical cybersecurity vulnerabilities that demand both accurate and interpretable intrusion detection solutions. Existing deep learning-based intrusion detection systems (IDS) achieve high detection accuracy but lack inherent explainability, limiting their clinical adoption under regulatory frameworks such as GDPR and FDA guidelines. This paper presents MedDefender-MHAN, an explainable multi-head attention network specifically designed for healthcare IoT threat detection. The proposed framework introduces a novel dual-stream architecture that combines convolutional neural networks for local spatial feature extraction with transformer-based encoders for long-range temporal dependency modeling. Unlike existing approaches that apply explainability as a post-hoc process, MedDefender-MHAN embeds interpretability directly into the multi-head attention mechanism, enabling real-time gradient-weighted explanation generation without external XAI pipelines. Evaluated on CICIDS2017 and TON_IoT benchmark datasets, MedDefender-MHAN achieves detection accuracies of 99.47% and 98.92% respectively, with sub-3ms inference latency and a throughput of 435 samples per second. Explainability evaluation demonstrates 94.6% alignment with expert-annotated attack signatures and 91.9% temporal accuracy, outperforming post-hoc methods such as SHAP and Integrated Gradients. These results confirm that MedDefender-MHAN provides a clinically viable, regulatory-compliant security solution for real-world healthcare IoT infrastructure. The proposed framework addresses the dual imperatives of methodological transparency and clinical impact, directly responding to the growing need for trustworthy AI-driven security solutions in regulated healthcare IoMT environments.
20 Apr 2026
Read appraisal →JMIR medical informatics
Quality of Clinical Notes Created by Ambient Listening Generative AI: Pragmatic Prospective Pilot Study
BACKGROUND: Physicians routinely document specifics of patient encounters in clinic visit notes, a critical but potentially time-consuming task. Ambient listening artificial intelligence (AI) technology is being integrated into clinical workflows to reduce documentation burden by creating draft visit notes. While this technology is promising, it is not perfect, and the potential for patient harm needs to be understood and mitigated. We developed and piloted an efficient, standardized approach to evaluating AI-generated notes for safety concerns in ambulatory care visits. OBJECTIVE: The objective of this quality improvement project was to develop and pilot an efficient, standardized, and scalable approach to evaluating AI-generated notes for safety concerns in ambulatory care visits. METHODS: During a 2-month pilot (July to August 2024), 31 physicians across multiple specialties used an ambient listening AI scribe to assist with the creation of 7545 clinic notes. A novel survey instrument was developed to assess note quality, focusing on 4 error types: accidental inclusions, accidental omissions, hallucinations, and bias. Physicians evaluated 356 (4.7%) AI-generated notes. Where an error was present, physicians rated its severity based on its potential to cause patient harm if it was not corrected, on a 0 to 5 scale. Additionally, a vendor-reported metric on the percentage of note content edited by physicians was analyzed. RESULTS: Of the 356 evaluated notes, accidental omissions were the most frequent error (n=64, 18%), followed by hallucinations (n=41, 11.5%), and accidental inclusions (n=33, 9.3%). Bias was rare (n=4, 1.1%). Most (119/142, 83.8%) errors were rated as mild to moderate (severity 1-3), with only 19 (5.3%) notes containing errors rated as posing serious or imminent risk (severity 4-5). Editing metrics across all AI-created notes showed a median of 9.0% (IQR 2.5%-21.9%) of AI-generated words were changed, with 14.9% (143/960) of notes left entirely unedited. Physician editing practices varied widely, with average percentages of AI-generated words changed ranging from 1.9% to 69.3% (median 9.0%, IQR 2.5%-21.9%). CONCLUSIONS: AI-generated clinical notes were generally of high quality, with 94.7% (337/356) free from significant errors. However, because a small number contained errors that carried the risk of serious harm if not corrected, careful clinician review of notes remains imperative. Prior to deploying an AI scribe, organizations should pilot the technology and include an efficient review process to understand the nature and type of errors common at their organization. This pilot provides a scalable model for other health systems seeking to implement AI scribe technology responsibly.
18 Apr 2026
Read appraisal →PloS one
EAC-Agent: A deep learning framework for multimodal emotion-aware conversational agent with contextual response generation
Conversational agents, commonly referred to as chatbots, have become an integral part of various important applications, such as customer support and virtual assistants in enterprise-level solutions. Despite their widespread use, current solutions predominantly rely on a single modality input, primarily text, which limits their ability to fully understand the correct emotions of users and respond accordingly. The increasing demand for multimodal inputs in conversation such as text, audio, and videos, highlights the need for a more comprehensive approach. Existing conversational agents face significant challenges in generating emotionally aware responses, as they lack the ability to effectively handle emotion embeddings, leading to limitations in emotional accuracy and contextual appropriateness. Focusing on these challenges, this research work leads to a novel multimodal approach which incorporates features from text, audio, and visual. The EAC-Agent proposed a sequence-to-sequence model with transformer, along with pre-trained embeddings such as GloVe. The self and cross-modal attention on text, audio and visual is used to generate a more emotionally intelligent response. EAC-Agent is validated through comparison with existing techniques on two benchmark datasets. The obtained results demonstrate superior performance in emotion classification and response generation. The proposed model achieves an accuracy of 76.27% on IEMOCAP and 67.57% on MELD for emotion recognition from multimodal user inputs. In addition, the emotion-aware response generation module shows clear improvements, with perplexity values of 39.01 and 42.30, BLEU scores of 0.31 and 0.30, and ROUGE-L scores of 0.45 and 0.44 on IEMOCAP and MELD, respectively. EAC-Agent demonstrates clear superiority over existing models and holds great promise for applications in customer service, healthcare, and other areas requiring empathetic and contextually appropriate interactions.
18 Apr 2026
Read appraisal →Philosophical transactions of the Royal Society of London. Series B, Biological sciences
Crowdsourced versus large language models forecasting: evidence for the accuracy-correlation effect
Over the past quarter century, crowdsourced forecasting has largely outperformed individual forecasters. Today, large language models (LLMs), aggregating human knowledge at scale, constitute a new form of collective intelligence (CI). A central question is how LLM predictive accuracy associates with human-AI correlation, and whether this relationship exceeds what would be expected if both merely track the same underlying truth. We investigate this through the accuracy-correlation effect (ACE), which posits that as algorithmic systems improve, they increasingly correlate with human predictions, potentially diminishing human value in hybrid ensembles. Using 76 model × prompt forecast sets from 16 LLMs on 580 resolved ForecastBench questions, we computed LLM accuracy and correlations with two human aggregates (superforecasters, general public), separately for databases (n = 526) and prediction markets (n = 54) questions. Linear mixed-effects models show a robust positive association between LLM accuracy and human-AI correlation that substantially exceeds independent-errors predictions. Correlations were lower for superforecasters than for the general public, and weaker for markets than for data questions. These results support ACE while indicating that increasing correlation reflects more than improved signal tracking alone, suggesting that simultaneous increases in accuracy and correlation may reduce optimal human weights in data-rich settings, while human judgement retains critical value in contextual reasoning scenarios. This article is part of the theme issue 'The evolution of collective intelligence'.
17 Apr 2026
Read appraisal →JMIR formative research
Development of a Contextualized, Research-Based Flemish Assessment Framework for Digital Care, Assistance, and Support: Delphi Study
BACKGROUND: The rapid evolution of digital technologies has transformed health, mental health, and social care, offering new modalities of digital care, assistance, and support through web-based platforms, mobile apps, extended reality, wearables, and artificial intelligence systems. Despite this proliferation, there is little consensus on what constitutes "high-quality" digital care. Challenges persist regarding data security, interoperability, accessibility, sustainability, and professional competence, whereas existing standards and regulations provide fragmented guidance. OBJECTIVE: This study aimed to develop a contextualized, consensus-based quality assessment framework for digital care, assistance, and support in Flanders, Belgium. For this purpose, perspectives across technology, organizational processes, and professional competencies were integrated. METHODS: The study used a multiphase design comprising (1) 10 expert interviews with Flemish government officials; (2) a narrative literature review of 303 peer-reviewed and gray literature sources; (3) a 3-round Delphi study with 50 experts across 5 domains (end users, facilitators, technology developers, deontology and ethics experts, and digital inclusion and media literacy experts); and (4) 4 complementary focus groups and 3 interviews with specialists in artificial intelligence, regulation, social work, mental health, and IT. The Delphi rounds gathered iterative feedback through open-ended elicitation, structured rating, and classification of quality criteria. Quantitative data were analyzed using descriptive statistics, whereas qualitative feedback was subjected to thematic analysis. RESULTS: A total of 50 experts participated in round 1, a total of 40 (80%) participated in round 2, and 27 (54%) participated in round 3. Round 1 generated 577 unique quality criteria, consolidated into 26 clusters organized under 3 pillars: technology, organization, and professional competencies. The relative importance across pillars was balanced (mean score 37.29, SD 12.38 for technology; 33.33, SD 10.39 for professional competencies; and 29.80, SD 10.45 for organizations). Accessibility, reliability, and safety ranked highest for the technology; vision, quality monitoring, and infrastructure ranked highest for organization; and support, digital competencies, and ethics ranked highest for professional competencies. The finalized framework included 112 criteria, of which 35 (31.3%) were designated as optional and 77 (68.8%) were designated as minimum requirements. Focus groups and interviews validated the framework's comprehensiveness and usability, emphasizing proportional implementation, user centrality, and alignment with European Union regulations. Stakeholders highlighted the need for tools, training, and governance mechanisms to ensure adoption and sustainability. CONCLUSIONS: This study produced a codeveloped, context-sensitive quality assessment framework that balances technological robustness, organizational readiness, and professional competence in digital care, assistance, and support. The framework can serve both as a quality safeguard and a developmental road map. Accompanying self-assessment and governance tools enhance practical applicability. Implementation success will depend on governmental support, resource allocation, and structured feedback loops. Future research should pilot the framework in real-world settings, assess its impact, and establish mechanisms for continuous updates to maintain relevance in a rapidly evolving digital landscape.
17 Apr 2026
Read appraisal →JMIR medical informatics
Classification of Cochrane Plain Language Summaries by Conclusiveness Using Transformer-Based Models and ChatGPT: Retrospective Observational Study
BACKGROUND: Cochrane plain language summaries (PLSs) aim to make systematic review findings more accessible to the general public. However, inconsistencies in how conclusions are presented may impact comprehension and decision-making. Classifying PLSs based on conclusiveness can improve clarity and facilitate informed health decisions. OBJECTIVE: This study aimed to develop and evaluate deep learning language models for the classification of PLSs according to 3 levels of conclusiveness (conclusive, inconclusive, and unclear) and to compare their performance with a general-purpose large language model (GPT-4o). METHODS: We used a publicly available dataset containing 4405 Cochrane PLSs of systematic reviews published until 2019, already classified by humans according to 9 categories of conclusiveness regarding the intervention's effectiveness or safety. We merged these categories into 3 classes based on the strength of conclusiveness: conclusive, inconclusive, and unclear. For the fine-tuning, we used Scientific Bidirectional Encoder Representations from Transformers (SciBERT), a pretrained language model trained on 1.14 million papers primarily from the health sciences, and Longformer, a transformer model designed specifically to process long documents. The script was developed using the Python programming language and the PyTorch framework. We computed evaluation metrics using the scikit-learn machine learning library and determined the area under the curve of the receiver operating characteristic (AUCROC) to measure the model performance in balancing sensitivity and specificity. We also analyzed a separate set of 213 PLSs and compared the predictions of our pretrained models with both manual verification and outputs generated by ChatGPT. RESULTS: The model based on SciBERT achieved a balanced accuracy of 56.6%. The AUCROC was 0.91 for "conclusive," 0.67 for "inconclusive," and 0.75 for "unclear" conclusiveness classes. The Longformer-based model had a balanced accuracy of 60.9%, with AUCROCs of 0.86 for "conclusive," 0.67 for "inconclusive," and 0.72 for "unclear" conclusiveness classes. Both models underperformed compared with ChatGPT, which demonstrated higher accuracy (74.2%), better precision and recall, and a higher Cohen κ (0.57). CONCLUSIONS: Fine-tuning 2 transformer-based language models showed mixed results in classifying Cochrane PLSs by conclusiveness, likely due to semantic overlap and subtle linguistic differences. Despite satisfactory internal test metrics, the fine-tuned models failed to generalize to newly published PLSs, where performance dropped to near-chance levels. These findings suggest that general-purpose large language models like GPT-4o may currently offer more reliable results for practical classification tasks in biomedical applications.
16 Apr 2026
Read appraisal →Clinica chimica acta; international journal of clinical chemistry
Artificial intelligence and machine learning applications in dialysis: Current applications, challenges, and future directions.
Artificial intelligence (AI) and machine learning (ML) applications have emerged as transformative technologies in nephrology, particularly in dialysis care. The availability of multimodal datasets including electronic health records, hemodialysis machine data, laboratory values, and imaging has enabled the development of sophisticated prognostic, diagnostic, and treatment support tools. This comprehensive review narratively examines current AI/ML applications in dialysis, evaluates their clinical performance, and identifies future research directions. We conducted a comprehensive literature search of PubMed, Web of Science, and other major databases from 2020 to 2025, focusing on peer-reviewed studies that employed AI/ML techniques in dialysis care. Studies were categorised by application domain and analysed for methodology, performance metrics, and clinical implications. Our analysis identified five major application domains: (1) prediction and prognosis, including intradialytic hypotension (IDH) prediction with AUROC values ranging 0.89-0.95, mortality prediction achieving C-indices up to 0.83, and hospitalisation risk assessment; (2) early detection of chronic kidney disease (CKD) progression and dialysis risk; (3) clinical decision support for anaemia management and treatment optimisation; (4) vascular access monitoring with AI-driven image analysis achieving AUROC ≈ 0.96; and (5) natural language processing (NLP) applications for symptom detection. Federated learning (FL) approaches are emerging to enable multi-centre collaboration while preserving data privacy. AI/ML technologies demonstrate significant promise in enhancing dialysis care through improved prediction accuracy, personalised treatment approaches, and clinical decision support. However, widespread clinical adoption remains limited due to challenges including data privacy concerns, model interpretability issues, regulatory complexity, and the need for diverse, representative datasets.
16 Apr 2026
Read appraisal →Journal of medical Internet research
A Health-Related Digital Ecological Momentary Assessment in Children (Aged 5-11 Years): Systematic Review
BACKGROUND: Digital ecological momentary assessment (EMA) collects data on experiences as they occur in daily life, capturing dynamic, context-sensitive experiences often missed by retrospective reporting. While EMA shows promise for pediatric health research, preadolescents have distinct socioemotional and cognitive characteristics likely to affect engagement. Existing reviews have not focused on the acceptability and feasibility of EMA protocols for this age group. OBJECTIVE: This review aimed to examine digital EMA protocols used with children aged 5-11 years across health domains, focusing on protocol characteristics, acceptability, and feasibility. We address 3 research questions (RQs)-RQ1: What are the characteristics of these protocols? RQ2: What is the feasibility and acceptability of these protocols? RQ3: What are the characteristics of high and low response rate protocols? METHODS: We searched 10 databases (CINAHL, Embase, ACM Digital Library, IEEE Xplore, Cochrane Library, PsycINFO, Web of Science, PubMed, Scopus, and MEDLINE) for peer-reviewed studies published up to October 2025. Eligible studies used EMA with children aged 5-11 years to collect health data via digital devices. Two researchers independently screened studies (SC and LT); one (SC) conducted quality assessment and data extraction. Findings were narratively synthesized. RESULTS: We identified 17 protocols across 37 studies. Most targeted nonclinical populations, used handheld devices, spanned 3-28 days, and applied interval-contingent prompting (RQ1). Response rates were available or calculable for 15 of 17 protocols, ranging from 48% to 92% (RQ2). Six protocols reported response rates of ≥80%. However, key data required for pooling (eg, raw counts for planned vs completed prompts) were missing or selectively reported. This contributed to 13 of 17 protocols being rated at critical risk of bias (ROBINS-I, v2). As a result, the strength of evidence was limited by poor reporting and high risk of bias. Facilitators included uncomplicated, engaging technology, reminders, and caregiver involvement. Barriers included device burden, restricted device access, difficulty with accurate reporting, stigma, limited device awareness, and insufficient caregiver support. High-response protocols (≥80%) often involved older children or clinical groups, ≥3-week duration, fixed schedules (≥20 items per prompt, 3 or 4 times per day), timing customization, and incentives (RQ3). CONCLUSIONS: This review provides the first systematic synthesis on preadolescents, offering insight into EMA protocol design beyond prior work treating children as a single group. By examining 17 EMA protocols, the review identifies gaps in developmental appropriateness and reporting quality, highlighting where the evidence may differ from adolescent and adult EMA research. The results suggest that digital EMA for preadolescents requires greater focus on child-centered design to increase acceptability and adherence, alongside improved reporting standards, so protocols can be meaningfully compared. With these advances, EMA could be more effectively integrated into pediatric health monitoring, tailored to the needs of different age groups.
16 Apr 2026
Read appraisal →International journal of medical informatics
Prioritising digital health technologies in Australian community pharmacy: a delphi study identifying barriers, enablers, and policy implications for implementation
OBJECTIVE: To determine priority digital health technologies for Australian community pharmacies and identify the key barriers, enablers, and policy/funding factors that can inform future implementation planning through expert consensus. METHODS: A two-round Delphi study was conducted with 31 experts representing pharmacy, academia, policy, and digital health. In Round 1, participants identified priority technologies, barriers, and enablers. In Round 2, 27 participants ranked five technologies, nine policy options, and six financial models. Consensus was assessed using descriptive statistics and interquartile ranges (IQRs). RESULTS: E-prescriptions and My Health Record (MyHR) were ranked as top priorities (mean = 1.70 and 2.22; IQR ≤ 1.0). Key barriers included financial constraints, interoperability issues, and digital literacy gaps. Telehealth incentives received the strongest agreement among participants, while reimbursement-based funding and government support were rated as the most supportive financial models for implementation. Broader enablers, such as a national medicine repository and stronger cross-disciplinary collaboration, were also endorsed. CONCLUSION: Digital health adoption in community pharmacy requires prioritisation of core technologies, improved system integration, workforce training, and practical funding mechanisms. These findings offer guidance for policymakers, pharmacy leaders, and digital health stakeholders aiming to embed digital tools more consistently and effectively into pharmacy practice.
15 Apr 2026
Read appraisal →PloS one
Unsupervised industrial image defect detection based on autoencoder and GANs
In the process of intelligentization in modern manufacturing, especially in industrial fields such as automobile manufacturing, semiconductor production, and electronic product assembly, product quality control is crucial. Traditional defect detection methods face problems such as supervised learning methods relying on a large amount of labeled data, weak generalization ability, and high cost. In the process of intelligent manufacturing, industrial product quality control is a key link to ensure production safety and product consistency. Especially in typical industrial scenarios such as automobile manufacturing, semiconductor production, and electronic product assembly, traditional defect detection methods are difficult to meet the needs of actual production lines for high-precision and high-efficiency detection due to their reliance on a large amount of labeled data, weak generalization ability, and high cost. To solve the problem of defect detection in small and zero sample scenarios in these industries, improve detection sensitivity and localization accuracy, and enhance the model's generalization ability to unknown defects, a unsupervised industrial image defect detection method based on autoencoder and Generative Adversarial Networks (GANs) is proposed. This study constructs an Multi-level Deep feature Adaptive fusion AutoEncoder (MDAAE) module, extracts multi-scale features through Pre-Trained Convolutional Neural Backbone Networks (PTCNBN), introduces Attention Mechanism (AM) to dynamically calculate feature weights, and achieves feature fusion and reconstruction. Meanwhile, the self-AM is fused to improve the GANs. The self-attention module efficiently captures long-range dependencies and generates an adversarial network training objective function to optimize the generator's unsaturated loss. The results showed that the Area Under the Curve (AUC) of the Recall-Precision curve of the research method reached 93.6 ± 0.5%, and its F1-Score value exceeded 0.890 ± 0.003 in defect types such as scratches and dents. In practical applications, the inference time of the research method remained stable at 3.0 ± 0.2 GB of central processor memory, and the Fréchet Inception Distance (FID) value fully converged at the 2230th iteration, with a stable FID value of 3. The false alarm rate was only 8.6 ± 0.7% under strong light conditions. The proposed UIIDD method based on autoencoder and GANs had good robustness, generalization ability, reliability, and efficiency. This study effectively solves the problems of poor robustness, weak generalization ability, and high cost of traditional defect detection methods.
12 Apr 2026
Read appraisal →JMIR human factors
Predicting Electronic Health Record Usability: Scoping Review of Adoption Models, Metrics, and Future Directions
BACKGROUND: Electronic health records (EHRs) play an essential role in modern health care, enabling data sharing and improving patient safety; however, even though vendors must adhere to International Organization for Standardization-related usability standards for EHR certification, persistent usability issues continue to undermine efficiency, contribute to clinician burden, and increase the risk of preventable errors. OBJECTIVE: This scoping review synthesizes existing research on EHR adoption and usability, emphasizing theoretical models, measurement approaches, factors, and analytic methods used to assess or predict usability. We identify gaps and opportunities for integrating predictive analytics and artificial intelligence (AI) to advance research and improve the usability of EHRs. METHODS: Following Joanna Briggs Institute and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines, we systematically searched MEDLINE, Web of Science, IEEE Xplore, and Scopus library databases for studies published between January 1, 2009, and April 9, 2025. Inclusion criteria focused on empirical research using predictive methods or models related to EHR usability. Data were charted and synthesized thematically. RESULTS: Of the 2323 screened papers, 47 studies met inclusion criteria. Most research examined or predicted EHR adoption (not usability) using dominant frameworks, such as the technology acceptance model, unified theory of acceptance and use of technology, and the information system success model, which comprised usability. Factors related to usability-particularly perceived usefulness, perceived ease of use, effort expectancy, and facilitating conditions via the EHR adoption models-appeared frequently. Regression-based methods and structural equation modeling were the most common analytic techniques. No studies applied predictive modeling or AI to predict EHR usability. CONCLUSIONS: The focus of this study on the prediction of EHR usability and adoption for the past 15 years is distinctive in the literature. It extends prior usability reviews (mostly focusing on adoption, not prediction of usability). Predictive modeling for EHR usability remains underdeveloped throughout 2009 to 2025. Dominant frameworks in the EHR literature continue to prioritize predicting adoption over operational usability. These models rely heavily on self-reported, cross-sectional measures captured at a single postimplementation time point, embedding systematic bias and obscuring longitudinal usability dynamics. Despite the application of increasingly sophisticated predictive techniques-primarily variants of regression and structural equation modeling-usability has remained analytically subordinate to adoption and acceptance constructs for more than 15 years. As a result, widely used models, such as the technology acceptance model and unified theory of acceptance and use of technology, position usability merely as an antecedent to intention or use, rather than as an independent, system-level property that can be empirically measured, modeled, and predicted. Therefore, there is substantial opportunity to integrate predictive analytics, AI, and longitudinal usability measures to build dynamic models.
10 Apr 2026
Read appraisal →JMIR human factors
Chatbot Usability Scale in Chinese Users: Cross-Cultural Adaptation and Validation Study
BACKGROUND: Chatbots are increasingly deployed across various domains; however, systematic evaluation of their usability remains limited, particularly in non-Western contexts. The 11-item Chatbot Usability Scale (BUS-11), a multidimensional instrument grounded in human-computer interaction theory, has demonstrated strong psychometric properties in prior studies; however, a validated Chinese version does not exist, despite China being one of the largest chatbot markets. OBJECTIVE: This study aimed to translate, culturally adapt, and validate the BUS-11 for Chinese users. METHODS: Following established cross-cultural adaptation procedures, the scale was forward- and back-translated, reviewed by an expert committee, and pilot-tested for clarity and feasibility. A main validation study was then conducted with 214 participants, who completed 438 evaluations of the chatbots across 10 widely used systems. RESULTS: Psychometric analyses demonstrated excellent scale-level content validity index (ie, 0.92), strong internal consistency (Cronbach α=0.92), and a clear 3-factor structure (Accessibility, Interaction Process Quality, and Information Quality), explaining 56.1% of the variance. Meanwhile, privacy or security and response time were retained as single-item indicators. The Chinese BUS-11 was concise, user-friendly, and psychometrically robust. CONCLUSIONS: This work fills a critical gap by providing the first validated instrument for assessing chatbot usability in Chinese contexts, enabling reliable cross-cultural comparisons and supporting both research and practical design evaluation in human-computer interaction.
7 Apr 2026
Read appraisal →PloS one
Identifying researcher characteristics driving growth in Japanese University-originated deep-tech startups: A machine learning approach
Active involvement of researchers with strong outputs is essential for growth of deep-tech university-originated startups. We used the University-Originated Venture Database to link total funding amount with researchers' publication metrics and KAKENHI grant records. Classification models were constructed to distinguish growing and non-growing startups, followed by feature-importance and breakdown-tree analyses to interpret which researcher attributes drive growth. We then applied these defined characteristics to evaluate researcher profiles at top Japanese universities to assess their entrepreneurial engagement. This process revealed a disconnect between high-quality research achievements and actual startup activity. These insights underpin an identification framework that can guide investment and policy decisions in Japan's deep-tech ecosystem.
5 Apr 2026
Read appraisal →Pediatric radiology
Data mining in pediatric radiology in the era of artificial intelligence
Data mining is the systematic process of extracting useful knowledge from large multimodal datasets and is increasingly enabled by artificial intelligence (AI) methods. Pediatric radiology is a natural field for data mining because multimodal data sources, including images, reports, metadata, and electronic health records, together capture rich information on anatomy, disease, treatment, and outcomes. In the current era, the boundaries between data mining and AI are increasingly blurred. AI assists in key steps of the mining workflow through automated labeling, information extraction, and representation learning, while data mining provides the high-quality curated datasets that underpin model performance, generalizability, and safety. This review, therefore, examines both domains together, emphasizing their interdependence in the pediatric context. We describe core concepts and workflows of data mining in pediatric radiology, including data collection, linkage, annotation, analysis, validation, and governance, and outline how modern AI tools such as deep learning, large language models, multimodal fusion, and federated learning support advanced pattern discovery across limited and heterogeneous pediatric datasets. We summarize current and emerging clinical applications across diagnosis, prognosis, radiation dose monitoring, operational analytics, reporting safety nets, and continual learning. We then discuss current challenges related to data quality and standardization, ethics, regulation, workflow integration, resource disparities, sustainability, and explainability. Finally, we highlight future perspectives, including synthetic data generation, foundation models, structured reporting, and pediatric-focused ethical frameworks that aim to enable safe, transparent, and equitable integration of AI-driven data mining to improve outcomes in children.
4 Apr 2026
Read appraisal →Current opinion in anaesthesiology
Data integration and systems interoperability: the prerequisite for artificial intelligence in anesthesiology
PURPOSE OF REVIEW: Anesthesiology generates large volumes of heterogeneous perioperative data, including high-resolution physiological signals, clinical documentation, and device-generated information. Despite this richness, the clinical deployment of artificial intelligence systems remains limited. This review examines how limitations in data integration and systems interoperability constrain the translation of artificial intelligence into routine practice. RECENT FINDINGS: Recent literature indicates that the principal barrier to artificial intelligence adoption in anesthesiology is not algorithmic performance but inadequate data integration. Data are distributed across anesthesia information management systems, electronic health records, and multiple medical devices, with variable data models, inconsistent semantic representations, and limited temporal synchronization. These systems were primarily designed for documentation, billing, and medicolegal purposes, rather than real-time analytics or secondary data use. In contrast, intensive care medicine has benefited from early investments in shared databases, which have facilitated reproducible artificial intelligence research and validation. Although emerging standards such as HL7 Fast Healthcare Interoperability Resources and common data models provide technical frameworks for interoperability, their implementation in anesthesiology remains partial. SUMMARY: The effective deployment of artificial intelligence in anesthesiology depends on the development of interoperable, high-quality data infrastructures. Establishing standardized data models, semantic harmonization, temporal alignment, and robust data governance is a prerequisite for scalable, trustworthy artificial intelligence-enabled perioperative care.
4 Apr 2026
Read appraisal →Clinical nutrition ESPEN
Artificial intelligence in clinical nutrition. A narrative review.
The rapid integration of Artificial Intelligence (AI) into healthcare, particularly clinical nutrition, holds transformative potential for enhanced diagnostics, risk prediction, and personalized therapeutic support. However, many clinicians lack sufficient understanding of AI principles, capabilities, and limitations, which may hinder adoption, lead to inappropriate use, or result in missed opportunities to enhance patient care. This narrative review aims to provide an accessible overview of foundational AI concepts, such as machine learning, deep learning, and large language models, and their practical applications in clinical nutrition. While AI offers immense promise in advancing nutritional care, its successful implementation requires clinicians to be adequately prepared to engage with these technologies. Education programs tailored to healthcare professionals, interdisciplinary collaboration between AI experts and clinicians, and robust ethical oversight are critical to ensure responsible and effective integration. By equipping clinicians with the necessary knowledge and tools, AI can serve as a powerful ally in delivering personalized and patient-centered nutritional care while maintaining human expertise at the forefront of decision-making.
2 Apr 2026
Read appraisal →Am J Perinatol
Application of Generative AI to Enhance Obstetrics and Gynecology Research
The rapid evolution of large-language models such as ChatGPT, Claude, and Gemini is reshaping the methodological landscape of obstetrics and gynecology (OBGYN) research. This narrative review provides a comprehensive account of generative AI capabilities, key use cases, and recommended safeguards for investigators. First, generative AI expedites hypothesis generation, enabling researchers to interrogate vast corpora and surface plausible, overlooked questions. Second, it streamlines systematic reviews by composing optimized search strings, screening titles and abstracts, and identifying full-text discrepancies. Third, AI assistants can draft reproducible analytic code, perform preliminary descriptive or inferential analyses, and create publication-ready tables and figures. Fourth, the models support scholarly writing by suggesting journal-specific headings, refining prose, harmonizing references, and translating technical content for multidisciplinary audiences. Fifth, they augment peer-review and editorial workflows by delivering evidence-focused critiques. In educational settings, these models can create adaptive curricula and interactive simulations for trainees, fostering digital literacy and evidence-based practice early in professional development among clinicians. Integration into clinical decision-support pipelines is also foreseeable, warranting proactive governance. Notwithstanding these opportunities, responsible use demands vigilant oversight. Large-language models occasionally fabricate citations or misinterpret domain-specific data ("hallucinations"), potentially propagating misinformation. Outputs are highly prompt-dependent, creating a reliance on informed prompt engineering that may disadvantage less technical clinicians. Moreover, uploading protected health information or copyrighted text raises privacy, security, and intellectual property concerns. We outline best-practice recommendations: maintain human verification of all AI-generated content; cross-validate references with primary databases; employ privacy-preserving, on-premises deployments for sensitive data; document prompts for reproducibility; and disclose AI involvement transparently. In summary, generative AI offers a powerful adjunct for OBGYN scientists by accelerating topic formulation, evidence synthesis, data analysis, manuscript preparation, and peer review. When coupled with rigorous oversight and ethical safeguards, these tools can enhance productivity without compromising scientific integrity. Future studies should quantify accuracy, bias, and downstream patient impact. · Generative AI supports various research stages in OBGYN, such as hypothesis generation, systematic review assistance, data analysis, and scientific writing, demonstrating its potential to streamline research workflows and improve research efficiency.. · Generative AI has notable limitations, including the risk of generating inaccurate references ("hallucinations") and the need for careful supervision.. · Effective usage requires skill in prompt engineering, posing a challenge for those without technical expertise.. · Utilizing generative AI in sensitive fields like OBGYN raises privacy, security, and ethical concerns..
1 Dec 2025
Read appraisal →