Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 56 appraisals
Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association
Artificial Intelligence Tools for Gastrointestinal Research: A Practical Guide
Artificial intelligence (AI) tools, including large language models (LLMs) such as ChatGPT, Claude, and Gemini, are increasingly used in biomedical research. However, practical guidance for integrating these tools into the manuscript writing workflow remains limited. This narrative review provides a structured, step-by-step guide for gastrointestinal (GI) researchers and clinicians seeking to use AI responsibly across the stages of manuscript preparation.
2 Aug 2026
Read appraisal →Langenbeck's archives of surgery
The role of large language models in the writing of surgical reviews: fact or fantasy?
BACKGROUND: Amidst the current enthusiasm concerning artificial intelligence and its possible application in the composition of different kinds of scientific and non-scientific written documents, we evaluated the usage of artificial intelligence for writing surgical short reviews. METHODS: In order to assess the formal and content quality of AI-generated texts compared to human written texts, ten AI-based text generators (five chatbots and five content creators) and four surgeons in training received the same prompt for a short scientific article on a liver surgery theme. All texts were anonymized and subsequently evaluated by three experienced liver surgeons based on a pre-defined scoring scheme, as well as for quality of references and readability according to readability indices. Furthermore, all texts were tested for plagiarism using PlagScan. RESULTS: Overall percentage of correct assessment for AI/non-AI generation by experienced surgeons lay at 78.57%. Human written text had a mean word count of 1054 versus 874 in AI-generated text, with a higher mean Flesh Reading Ease Score (FRE, 26.2 ± 5.1 versus 17.7 ± 6.1). References were PubMed-listed in 100% for human versus 46% for AI-generated text, with only one non-human text reaching 100% formally correct citation of references. PlagScan found 6.4%±1.3 mean resemblance to existing texts for human versus 7.6%±4.5 for AI-generated text. DISCUSSION: Overall, AI could already mislead experienced scientific surgeons in 26.7% of cases into believing it to be human. However, formal requirements, especially considering referencing, are still in great need of improvement with only one of AI-generated articles fulfilling our quality requirements.
4 July 2026
Read appraisal →Journal of the American College of Clinical Pharmacy : JACCP
Research and Scholarly Methods: Using Artificial Intelligence
Artificial intelligence (AI) is quickly changing health care and expanding research opportunities, including in pharmacy. AI can offer effective solutions for numerous challenges in pharmacy research, such as persistent gaps in evidence-based prescribing, slow drug discovery-to-adoption timelines, complex data environments, and the need for more efficient methods, including qualitative methods. AI tools-ranging from machine learning models to large language models (LLMs)-offer new capabilities for data analysis, prediction, and content generation. Pharmacy researchers can harness these technologies to address persistent gaps in medication management, optimize patient safety and outcomes, and streamline many research workflows. However, successful integration requires understanding the spectrum of AI capabilities, appropriate use cases, and potential limitations. AI must be tailored to the research question, data, and institutional resources. This paper will guide pharmacy researchers in identifying use cases for AI in their research, selecting and applying AI tools, mitigating risks, providing standards for reporting the use of AI in research, and maximizing benefits for scientific discovery and patient outcomes.
1 July 2026
Read appraisal →Phonetica
Gradient boundaries through confidence intervals for forced alignment estimates using model ensembles
Forced alignment is a common tool to align audio with orthographic and phonetic transcriptions. Most forced alignment tools provide only point-estimates of boundaries. The present project introduces a method of producing gradient boundaries by deriving confidence intervals using neural network ensembles. Ten different segment classifier neural networks were previously trained, and the alignment process is repeated with each classifier. The ensemble is then used to place the point-estimate of a boundary at the median of the boundaries in the ensemble, and the gradient range is placed using a 97.85 % confidence interval around the median constructed using order statistics. Gradient boundaries are taken here as a more realistic representation of how segments transition into each other. Moreover, the range indicates the model uncertainty in the boundary placement, facilitating tasks like finding boundaries that should be reviewed. As a bonus, on the Buckeye and TIMIT corpora, the ensemble boundaries show a slight overall improvement over using just a single model. The gradient boundaries can be emitted during alignment as JSON files and a main table for programmatic and statistical analysis. For familiarity, they are also output as Praat TextGrids using a point tier to represent the edges of the boundary regions.
27 June 2026
Read appraisal →Journal of Korean medical science
Evaluating Large Language Models for Post-Publication Promotion: A Blinded Comparative Study of Social Media Posts in Public Health
BACKGROUND: Social media platforms such as X (formerly Twitter) are increasingly used by journals, authors, and institutions to promote newly published research. Well-designed posts can enhance visibility, accelerate knowledge translation, and increase altmetric attention. However, creating accurate and policy-compliant content is time-intensive. Large language models (LLMs) offer a potential solution, yet systematic evaluations of their performance in post-publication promotion remain limited. METHODS: We conducted a blinded, crossed, offline evaluation of four LLMs: GPT-5 (OpenAI), Gemini 2.5 Pro (Google DeepMind), Grok-3 (xAI), and Perplexity Pro (Perplexity AI), tasked with generating X-style posts (≤ 260 characters) for 36 open access articles from The Lancet Public Health, The Lancet Planetary Health, and Annual Review of Public Health. Posts were generated using a standardized system and user prompt. A single blinded rater scored outputs using a five-domain rubric (factual accuracy, clarity, policy compliance, call-to-action quality, structure/metadata; maximum score 10). Secondary measures included character count, hashtag use, and readability (Flesch-Kincaid Grade Level). General linear models with Bonferroni-adjusted post hoc tests and non-parametric analyses were applied. RESULTS: All four models achieved perfect factual accuracy and no policy violations. Mean total quality scores differed significantly by model, P < 0.001. GPT-5 (9.60) and Perplexity Pro (9.60) performed best, followed by Gemini 2.5 Pro (9.47), while Grok-3 scored lower (8.80). Domain analyses showed Grok-3 underperformed in call-to-action quality (1.40 vs. ≥1.97 in other models, P < 0.001) and produced significantly shorter posts (median 194 characters, P < 0.001). Perplexity Pro scored highest for policy compliance, while GPT-5 and Gemini 2.5 Pro achieved superior structural scores. Readability varied: GPT-5 8.9 (7.3-9.2) and Perplexity Pro 7.3 (6.5-8.8) generated more complex outputs, whereas Gemini 2.5 Pro 5.1 (4.8-6.5) and Grok-3 4.5 (3.6-6.3) produced more accessible posts. CONCLUSION: LLMs can reliably generate accurate and policy-compliant social media posts for research promotion, with differences in style and readability that may inform audience targeting. GPT-5, Gemini 2.5 Pro, and Perplexity Pro produced high-quality outputs, while Grok-3 underperformed across several domains. These findings highlight the potential of LLMs as scalable first-draft tools for post-publication promotion, capable of improving the reach and accessibility of scientific research. Careful model selection, tailored to audience and communication goals, together with human oversight, remains essential.
25 June 2026
Read appraisal →Systematic reviews
Using full agreement across multiple large language models for title-and-abstract screening in systematic reviews: a proof-of-concept
BACKGROUND: The exponential growth of scientific literature poses significant challenges for conducting systematic reviews, particularly in the labor-intensive title-and-abstract screening phase. This study examines the feasibility of using multiple large language models (LLMs) for title-and-abstract screening in systematic reviews. METHODS: We propose and evaluate a full-agreement approach using three commercially available LLMs from different model families (ChatGPT, Gemini, and Claude), in which automated classification decisions are only accepted when all three models assign the same label. This approach was examined across six datasets to assess its effectiveness. A structured workflow was developed to support implementation without requiring specialized technical expertise. Additionally, a stop criterion was introduced to ensure that LLM-based classification is only applied when predefined performance thresholds are met. RESULTS: Across the six datasets, approximately four in five abstracts received full agreement across models and could therefore be classified automatically. For this subset of abstracts, classification performance was consistently higher than that of previous automated approaches using LLMs, with statistically significant improvements in all prespecified performance metrics. CONCLUSIONS: A full agreement approach across three LLMs may offer a promising and conservative strategy for title-and-abstract screening in systematic reviews by automating concordant decisions while reserving human review for discordant cases. As a proof-of-concept, the present findings support this approach as a possible workflow for reducing screening burden in the face of continued growth in the scientific literature, although its broader generalizability remains to be established.
14 June 2026
Read appraisal →Nature communications
Adaptive Riemannian optimization for multi-scale diffeomorphic matching
Image matching is a fundamental task in quantitative biomedical and biological image analyses, enabling researchers to compare, integrate, and interpret imaging data across subjects, time points, modalities, and experimental conditions. Existing state-of-the-art registration methods are slow due to inefficient implementations and poor convergence rates because of the ill-conditioned nature of the optimization problem. Deep learning methods offer fast inference but require extensive training time, substantial inference memory, and fail to generalize across long-tailed distributions or diverse image modalities, necessitating costly retraining. We address these challenges by proposing FireANTs, a training-free, GPU-accelerated, multi-scale adaptive Riemannian optimization algorithm for fast and accurate dense diffeomorphic image matching. FireANTs more than doubles the speed of the community standard ANTs registration tool on a CPU, and is two orders of magnitude faster on a GPU. On the GPU, FireANTs performs competitively with deep learning methods on inference runtime while consuming up to 10 × less memory. FireANTs demonstrates robustness on a wide variety of matching problems across modalities, species, and organs, without any domain-specific training or tuning. Our framework allows hyperparameter grid search studies with less resources and time compared to traditional and deep learning registration algorithms alike.
12 June 2026
Read appraisal →European respiratory review : an official journal of the European Respiratory Society
Artificial intelligence-enabled histological analysis in pre-clinical respiratory disease models: a scoping review
Histological analysis is a cornerstone of pre-clinical respiratory disease research. It enables assessment of pathology, therapeutic effects, and mechanisms. However, conventional approaches rely on manual scoring, which is subjective, time-consuming, and difficult to scale. Artificial intelligence (AI), particularly deep learning, offers potential to automate histology workflows. To date, its use in pre-clinical respiratory models has not been synthesised.We conducted a scoping review following the Joanna Briggs Institute guidelines. We searched MEDLINE and Embase (inception - January 2025) for pre-clinical studies using AI to analyse histology in respiratory disease models. Screening, full-text review, and data extraction were performed in duplicate.Of 6271 studies screened, 29 met inclusion criteria. Most used murine models (76%) and investigated lung cancer (28%), pulmonary fibrosis (24%), or tuberculosis (17%). Haematoxylin and eosin was the most common stain (48%), with others targeting collagen or immune markers. AI tasks included image classification (n=20), segmentation (n=10), and object detection (n=4), predominantly using convolutional neural networks (69%). Pre-processing methods (e.g. stain normalisation) were common, but annotation and training practices were inconsistently reported. AI model performance was generally high (accuracy ≥90%; seven studies); however, validation metrics varied, and external validation was absent. Most studies used "black box" models, with minimal application of explainability techniques. Reproducibility measures, such as sharing datasets or code were rarely reported.AI tools are poised to transform histological analysis in pre-clinical respiratory research. The field will be able to further harness AI to automate pre-clinical respiratory histological analysis by addressing gaps that we have identified in validation, transparency, and standardisation.
12 June 2026
Read appraisal →Journal of medical Internet research
From Virtual Molecules to Clinical Trials: How AI Is Reshaping Preclinical Drug Discovery
Pharmaceutical development is typically a long and arduous process, but advances in artificial intelligence may help streamline each stage of this pipeline. In this News and Perspectives article, JMIR Correspondent Benedette Cuffari reports on recent innovations and their potential implications for drug discovery.
1 June 2026
Read appraisal →JMIR medical informatics
Blockchain Smart Contracts for Automating Clinical Trials: Systematic Review and Proposed System Architecture
BACKGROUND: Blockchain technologies have revolutionized the financial sector through their ability to generate immutable, cryptographically secure records. Clinical trials and health care data possess several synergies with those of the financial sector, specifically pertaining to the importance of tamper-resistant recording of processes. The evolution of blockchain to autonomously execute tasks contingent upon predefined contractual terms via smart contracts (SCs) allows a dynamic chain of interlinked events to unfold independently and in sequence, with time-stamped records. In recent years, mistrust in clinical trial data has grown significantly. Recording the entire clinical trial lifecycle from application, registration, recruitment to conduct, finance management, statistical analysis, and reporting in an immutable, cryptographically secure ledger with SC execution of trial processes could limit the potential for human intervention and tampering. This would produce a time-stamped record of all events within the trial lifecycle. Leveraging the capabilities of SCs could alleviate recruitment challenges and address ongoing concerns regarding data transparency, ownership, and integrity that currently undermine clinical trial processes. OBJECTIVE: This study aimed to review the existing literature on SC applications in clinical trials and propose a system architecture for using SCs to automate key processes throughout the clinical trial lifecycle. METHODS: A systematic search was conducted in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, identifying peer-reviewed studies and open-source repositories pertaining to the implementation of SCs in clinical trials. Data were extracted specific to the stage of the trial lifecycle described, SC architecture, and technical specifications for real-world implementation. Data were synthesized to propose an architecture for automating clinical trial processes within the lifecycle using SCs. RESULTS: A total of 144 records were screened; 10 studies met the inclusion criteria. Most implementations used private Ethereum-based networks (7/10, 70%). Reported applications included automated patient matching (5/10, 50%), consent management with dynamic permissioning (6/10, 60%), protocol enforcement and time-stamped audit logs (9/10, 90%), adverse event reporting (3/10, 30%), and financial or workflow automation (3/10, 30%). SC-based recruitment systems demonstrated rapid matching performance (eg, 6000 simulated patients matched in 2.13 s). However, all included systems were prototypes or simulations, and none were tested in real-world regulatory settings. Scalability, interoperability limitations, regulatory ambiguity (eg, General Data Protection Regulation right-to-erasure conflicts), and high infrastructural complexity were common gaps noted across studies. CONCLUSIONS: Current evidence suggests that SCs can enhance transparency, traceability, and automation throughout the clinical trial lifecycle. However, the literature remains dominated by simulation-based prototypes, primarily Ethereum-dependent architectures, and lacks analyses of cost-effectiveness, governance, and integration with institutional workflows. Future research should evaluate hybrid architectures, develop interoperability standards, and assess regulatory and ethical implications in real deployments.
30 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 →Behavior research methods
Automating data extraction in meta-research: A multi-model benchmark in network psychometrics papers
Manual data extraction in meta-research is often tedious, time-consuming, and error-prone. In this paper, we investigate whether the current generation of large language models (LLMs) can be used to extract accurate information from scientific papers. Across the meta-research literature, these tasks usually range from extracting verbatim information (e.g., the number of participants in a study, effect sizes, or whether a study is preregistered) to making subjective inferences. Using a publicly available dataset containing a wide range of metascientific variables from 43 network psychometrics papers, we tested five LLMs (Claude 4.6 Opus, Claude 4.5 Sonnet, Claude 4.5 Haiku, GPT-5.2, and GPT-5 mini). We used an automated API-based pipeline to extract variables from the documents. This approach allows batch processing of research papers. As such, it represents a more efficient and scalable way to extract metascientific data than the default chat interface. The extraction accuracy ranged from 79.6% to 91.3% across the models. The extraction performance was generally higher for more explicit, verbatim information and worse for variables that required more complicated inference. Furthermore, most models were able to convey uncertainty in more contentious cases. We provide a comparison of the accuracy and cost-effectiveness of the individual models and discuss the characteristics of variables that are and are not suitable for automatic coding. Furthermore, we describe some of the common pitfalls and best practices of automated LLM data extraction. The proposed procedure can substantially reduce the time and costs associated with conducting meta-research.
25 May 2026
Read appraisal →Postgraduate medical journal
Peer review in the age of artificial intelligence: a comparative study of human and AI-generated review reports
BACKGROUND: Peer review is central to maintaining scientific quality and helps editors make decisions. However, the volume of scientific publications continues to rise, placing pressure on the peer review system. With the rise of generative AI, its role in supporting peer review is gaining attention. This study aims to compare human-written and AI-generated peer review reports. METHODS: We analysed 398 peer review reports linked to 119 research articles published in BMJ Open in 2024. Publicly available reports and manuscripts were included. Editorials, corrections, and protocols were excluded. AI reports were generated using ChatGPT. All reports were anonymized and assessed by two independent reviewers. We conducted a hybrid thematic analysis. Frequencies of themes were calculated and compared by reviewer type. For quantitative comparison, we used the Mann-Whitney U test to assess differences in review quality scores and Fisher's exact test to compare the distribution of themes. All analyses were conducted using R software. RESULTS: Human reviewers gave more detailed and diverse comments. They addressed deeper issues like interpretation, originality, and applicability. AI reviews covered more sections but focused on routine or structural elements. AI outperformed slightly in format-related domains. Co-occurrence analysis showed human reviews linked diverse themes, while AI comments were structurally clustered. Shannon Index confirmed that human reviews were more thematically diverse. CONCLUSIONS: AI can support peer review by screening for basic errors. However, it lacks insight, critical judgment, and contextual awareness. Human input remains essential for meaningful review. Review-specific AI tools that preserve confidentiality are needed for future integration.
22 May 2026
Read appraisal →BMJ (Clinical research ed.)
Concordance between target trial emulation and randomised controlled trials: systematic review and meta-analysis
OBJECTIVES: To assess the concordance of results between paired target trial emulations and their benchmarking randomised controlled trials and to explore determinants of emulation success. DESIGN: Systematic review and meta-analysis. DATA SOURCES: Medline, EMBASE, Scopus, PsycINFO, and Web of Science (2010 to 2025). STUDY SELECTION: Observational studies that explicitly aimed to emulate target randomised controlled trials of a health intervention were included. DATA EXTRACTION: Data were extracted on study domain, target population, intervention, comparator, and main outcomes for each emulation and corresponding randomised controlled trial. MAIN OUTCOME MEASURES: For 21 predefined emulation design features, concordance was assessed using Pearson correlation coefficients, standardised difference agreement, and ratio of ratios, with subgroup and regression analyses of study characteristics and concordance. RESULTS: Among 107 target trial emulation-randomised controlled trial pairs, the Pearson correlation coefficient was 0.59 (95% confidence interval (CI) 0.45 to 0.70), standardised difference agreement was 79% (85/107), and summary ratio of ratios was 0.96 (95% CI 0.92 to 1.01; I2=36%). In 63 pairs with closer emulation of trial design, a higher level of agreement was observed, with a Pearson correlation coefficient of 0.83 (0.73 to 0.89) and standardised difference agreement of 87% (55/63). In subgroup analysis, trial emulations tended to systematically underestimate treatment effects from randomised controlled trials for specific outcomes related to venous thromboembolism and major adverse cardiovascular events, while overestimating those for respiratory outcomes, with pooled ratio of ratios of 0.76 (0.58 to 1.00; I2=40%), 0.91 (0.86 to 0.96; I2=32%), and 1.20 (1.03 to 1.40; I2=40%), respectively. Moreover, emulations based on claims data tended to underestimate the treatment effects (ratio of ratios 0.90, 0.82 to 0.99; I2=38%). In univariate regression analyses, poorer target trial emulation-randomised controlled trial concordance was associated with emulation designs with imbalanced baseline population characteristics, treatment started in hospital, and lower outcome emulation quality. CONCLUSIONS: Current target trial emulations achieve moderate concordance when replicating their corresponding randomised controlled trials. Concordance could be improved by better emulation design, including better baseline and outcome emulation and leveraging multisource linked databases. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42024619811.
20 May 2026
Read appraisal →Analytical chemistry
Adaptive Deep Learning for High-Fidelity Quantitative Single-Molecule Imaging
Accurate quantification of molecular dynamics in live cells is critical for elucidating receptor signaling and guiding therapeutic strategies. Yet current deep learning methods for fluorescence imaging often distort intensity and lack robustness under diverse imaging settings, limiting quantitative utility. We present an adaptive deep learning framework that constructs in situ training data sets from ongoing sequences and integrates a self-feedback algorithm to preserve absolute molecular intensities. This approach achieves ≥ 3.9-fold gains in signal-to-noise and 1.6-fold improvements in localization precision across organic dyes, quantum dots, and fluorescent proteins. By maintaining both spatial and intensity fidelity, it enhances super-resolution reconstruction, stepwise photobleaching analysis, and live-cell single-molecule tracking. Applied to programmed death-ligand 1 (PD-L1), it uncovers density-dependent clustering, with small-molecule inhibitors inducing dimerization and reduced mobility, while antibodies increase diffusivity. Combined treatments exert complementary effects on PD-L1 membrane organization. This adaptive and intensity-preserving framework provides a broadly applicable platform for high-precision quantitative bioimaging across modalities and experimental conditions.
20 May 2026
Read appraisal →Expert review of vaccines
Approaches to observational study designs and analytical options to evaluate the safety of multi-dose vaccines: a systematic review
INTRODUCTION: Observational studies require careful considerations when evaluating the safety of multidose vaccines. We reviewed design and analytical approaches in observational studies evaluating the safety of multidose vaccines in the post-licensure phase. METHODS: EMBASE, MEDLINE, Web of Science, and Scopus (2018-2022) were searched for hypothesis-testing studies evaluating the safety of multidose vaccines. Key features from frequently used designs were extracted. RESULTS: Among 123 eligible studies, cohort (46%) and self-controlled case series (SCCS)/self-controlled risk interval (SCRI) (40%) followed by case-control (12%) were the most common designs, and 15% of studies used multiple designs. Among cohort studies evaluating multiple doses, vaccination date (36%) and cohort entry with time-updated exposure status (32%) were frequent approaches used to define time zero. Twenty-eight percent of cohort studies did not report time zero; all but one evaluated COVID-19 vaccine effect on post-delivery and fertility-related outcomes. For SCCS/SCRI, 64% of studies accounted for event-dependent exposures, mainly by including pre-exposure periods (53%) and modified SCCS model (48%), while 20% employed multiple correction strategies. Among studies using multiple designs, 68% reached consistent conclusions. CONCLUSIONS: SCCS/SCRI and cohort designs dominate multidose vaccine safety studies. Clear reporting on time zero in pregnancy and fertility-related cohort studies, and on addressing event-dependent exposures in SCCS/SCRI studies is needed, along with guidance on interpreting results from multiple designs. Observational studies are important for assessing safety of vaccines especially for rare health outcomes that cannot be easily detected in randomized trials. Commonly used approaches include comparing these health outcomes between vaccinated and unvaccinated individuals or between different time points in relation to vaccination timing within individuals. However, both approaches have important design challenges. For example, comparison between vaccinated and unvaccinated individuals requires careful consideration of alignment of the time of starting to observe outcomes with the time when individuals become eligible and vaccination time for both comparison groups (time zero). Within-individual comparison requires that earlier occurrence of health outcomes does not influence whether someone receives later vaccine dose(s). These challenges become more complex to address when vaccines are given in multiple doses.In this review, we assessed how important design and analytical issues are handled and reported in published studies evaluating the safety of multidose vaccines and assessed agreement between different study designs when more than one is used.We found that 28% of studies comparing vaccinated and unvaccinated groups did not clearly report time alignment (time zero) mostly with studies on pregnancy and fertility outcomes. Similarly, about one-third of studies using within-individual comparison did not report checking and addressing the possibility that earlier outcomes could influence future vaccinations. When multiple designs were used, two-thirds of studies reached similar conclusions, although comparisons were mainly qualitative. Future work is needed to guide clear reporting of these issues to improve transparency in design and analysis and when comparing results across different designs.
19 May 2026
Read appraisal →Biological psychiatry
Overview of Challenges in Brain-Based Predictive Modeling: Toward Meaningful Predictive Insights
Predictive analytics based on machine learning (ML) and artificial intelligence is a powerful tool enabling precision psychiatry and providing insights into brain-behavior relationships. However, given the mixed results observed in the field so far, making meaningful progress requires careful consideration of several key challenges to ensure the validity of models and findings, including overfitting, confounding biases, site effect harmonization, and interpretability, among others. First, we highlight limitations of cross-validation, a ubiquitous ML strategy used to prevent overfitting and obtain generalization estimates, emphasizing the risk of performance inflation and the need for independent validation. Next, we introduce different types of so-called third variables that can influence the examination of a brain-behavioral relationship of interest in different ways, using causal inference principles. We emphasize the biasing impact of confounding variables on ML models and summarize common mitigation strategies. We then discuss site-specific effects in multisite datasets, reviewing different harmonization strategies to reduce unwanted variability and site-specific noise. Finally, we explore post hoc model interpretation methods to enhance model transparency while cautioning against misinterpretation. By integrating rigorous result validation, confounder control, and interpretability techniques, researchers can ensure that ML models produce more reliable and generalizable findings and avoid spurious associations.
18 May 2026
Read appraisal →Biological psychiatry
Multimodal Representation Learning for Parsing Biological Heterogeneity in Psychiatric Neuroimaging
For decades, psychiatric neuroimaging has searched for biomarkers of depression and other disorders, but they remain elusive in clinical practice. While the last 5 years have seen rapid progress, other large-scale correlative studies have found only small, unreliable links between brain measures and clinical symptoms. Growing evidence suggests that such limitations are not just about sample size but depend critically on how models represent data. This review traces a recent shift away from univariate methods to multivariate/multiview approaches that learn more effective representations of biological and symptom measures by flexibly learning multimodal latent representations. First, we review how linear multiview embedding methods have revealed reproducible biological depression subtypes but do not perform well in small samples or samples enriched for mild symptoms. Then, we consider newer work exploring more sophisticated representations for neuroimaging data, including deep-learning and graph-based representations, and multimodal extensions that uncover complex latent patterns that single-modality studies miss. Then, we review recent developments in foundation models, which, once trained on large corpora, can "transfer learn" readily to small clinical cohorts, potentially bringing the advantages of large-scale learning to small, privacy-limited data. Finally, we highlight emerging representation tools that treat the brain as a dynamic, stateful multivariate process. Taken together, these advances point to a future in which the value of neuroimaging will be determined not only by ever-larger sample sizes but also by data quality and by how well our algorithms capture the distributed, multimodal, and evolving nature of psychiatric disorders.
17 May 2026
Read appraisal →JMIR research protocols
A Natural Language Processing Framework for Structuring and Visualizing Clinical Trial Eligibility Criteria at Scale: Protocol for a Quantitative Study
BACKGROUND: Eligibility criteria are essential to clinical trial design, guiding recruitment, and ensuring patient safety and scientific rigor. However, criteria are often lengthy, heterogeneous, and inconsistently formatted, which hinders large-scale interpretation and slows patient-trial matching. Manual review is time-consuming and error-prone. Advances in natural language processing and large language models (LLMs) offer opportunities to standardize and analyze eligibility text at scale. OBJECTIVE: This study aims to develop and evaluate a scalable system that uses LLM-enabled natural language processing and unsupervised learning to identify, normalize, categorize, and visualize clinical trial eligibility criteria, with the goal of improving patient-trial matching and revealing domain-level trends. METHODS: We designed a three-part pipeline: (1) representation of eligibility text using embeddings, followed by clustering to group semantically similar criteria; (2) dual-layer zero-shot LLM summarization for concept normalization, refinement, and deduplication of cluster exemplars; and (3) an interactive, web-based visualization interface to explore criteria distributions and trends by disease domain and over time. The pipeline was applied to 53,872 oncology trials (breast, lung, and gastrointestinal cancer) indexed on ClinicalTrials.gov. Outputs include cluster labels, normalized criterion summaries, and per-domain frequency profiles. Feasibility was assessed via successful end-to-end processing and inspection of face validity for cluster coherence and domain-specific patterns. RESULTS: The system successfully processed all 53,872 trials and generated stable clusters of inclusion and exclusion concepts. The LLM summarization layers produced concise, nonredundant labels that improved the interpretability of clustered criteria. The visualization interface enabled rapid exploration of cross-trial patterns and temporal trends within breast, lung, and gastrointestinal oncology, facilitating identification of common inclusion requirements and potential barriers to enrollment. A public, open-source demonstration instance allows for interactive exploration of these clusters and summaries. Benchmarking through human validation on a random sample of eligibility criteria found the system to be 94% (470/500) accurate, reflecting its ability to consistently categorize criteria correctly in congruence with human judgment. CONCLUSIONS: A combined embeddings-clustering-LLM pipeline can standardize heterogeneous eligibility text and surface domain-level patterns at scale. This framework provides a foundation for accelerating patient-trial matching and informing future trial design. While the current implementation was evaluated on ClinicalTrials.gov oncology trials, the approach is readily generalizable to additional diseases and alternative modeling configurations.
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 →Molecular genetics and genomics : MGG
Comparison between a conventional tool and deep learning models for RNA velocity analysis of scRNA-Seq data
Single-cell RNA sequencing (scRNA-Seq) enables analysis of gene expression at single-cell resolution. RNA velocity analysis infers the temporal dynamics of transcriptional states from the relative abundances of spliced/unspliced mRNA quantified via scRNA-Seq. Classical RNA velocity approaches, such as scVelo, implement gene-specific kinetic modeling. Deep learning methods including DeepVelo, VeloVI, LatentVelo, SymVelo, and scTour are based on variational autoencoders (VAEs), which allow to enhance the robustness and accuracy by leveraging nonlinear latent representations. Here, we systematically evaluated the performance of deep learning RNA velocity tools by comparing with the scVelo dynamical model to access the possible advantages of VAE-base methods. For this purpose, public datasets (GSE149689 and GSE203233) were initially processed using a standard scRNA-Seq pipeline. Comparisons among results of selected velocity tools were conducted using cosine similarity of velocity vectors to assess directional concordance, and by mean squared error analysis of trajectory continuity for the deep learning models. Overall, VAE methods produced significant, richer, and more directionally coherent and consistent velocity fields than the classical model. Our findings indicate that deep learning models provide more consistent and biologically plausible cell-state trajectories, although at the expense of higher computational demands and reliance on accurate splicing quantification. Altogether, our results underscore the relevance of VAE-based frameworks to advance RNA velocity analysis while highlighting the need for careful preprocessing.
14 May 2026
Read appraisal →Behavior research methods
A stroke-level large-scale database of Chinese character handwriting and the OpenHandWrite_Toolbox for handwriting research
Understanding what linguistic components (e.g., phonological, semantic, and orthographic systems) modulate Chinese handwriting at the character, radical, and stroke levels remains an important yet understudied topic. Additionally, there is a lack of comprehensive tools for capturing and batch-processing fine-grained handwriting data. To address these issues, we constructed a large-scale handwriting database in which 42 Chinese speakers each handwrote 1200 characters in a handwriting-to-dictation task. Additionally, we enhanced the existing handwriting package and provided comprehensive documentation for the upgraded OpenHandWrite_Toolbox. This toolbox allows users to easily modify the experimental design, capture the stroke-level handwriting trajectory, and batch-process handwriting measurements (e.g., latency and duration). Analysis of our large-scale database indicates that orthographic predictors affect handwriting preparation and execution at the character, radical, and stroke levels. Phonological factors influence handwriting at character and radical levels. Importantly, these lexical effects demonstrate hierarchical attenuation: they were most pronounced at the character level, followed by the radical, and weakest at the stroke level. These findings demonstrate that handwriting preparation and execution at the radical and stroke levels are closely intertwined with linguistic components. This database and toolbox offer valuable resources for future psycholinguistic and neurolinguistic research on the handwriting of characters and sub-characters across different languages.
14 May 2026
Read appraisal →Biology of sex differences
Importance of integrating biological sex and age analyses in health research
Research findings in human, animal and cell populations may be influenced by biological sex and age. The traditional statistical approach to sex and/or age differences that are discovered in research data is to consider them as confounders that interfere with the ability to make accurate estimations and thus need to be controlled for. This article provides examples of how combining sexes or controlling for sex and/or age may lead to inaccurate findings and overlook important insights. Rather than treating these variables as confounding factors, they should be viewed as variables of importance to the research question. We contend that data should be analyzed according to sex/age as the primary analysis, while controlling for sex and age should be conducted as a secondary analysis. This topic is increasingly important as data are made publicly available and deep learning models/artificial intelligence are used to analyze large volumes of data. Mechanistic insights will continue to be lost if prompt action is not taken to report data according to sex and age and to analyze data by sex and age in health research.
14 May 2026
Read appraisal →Neuroinformatics
Instance-Aware Pseudo-Labeling and Class-Focused Contrastive Learning for Weakly Supervised Domain Adaptive Segmentation of Electron Microscopy
Annotation-efficient segmentation of the numerous mitochondria instances from various electron microscopy (EM) images is highly valuable for biological and neuroscience research. Although unsupervised domain adaptation (UDA) methods can help mitigate domain shifts and reduce the high costs of annotating each domain, they typically have relatively low performance in practical applications. Thus, we investigate weakly supervised domain adaptation (WDA) that utilizes additional sparse point labels on the target domain, which require minimal annotation effort and minimal expert knowledge. To take full use of the incomplete and imprecise point annotations, we introduce a multitask learning framework that jointly conducts segmentation and center detection with a novel cross-teaching mechanism and class-focused cross-domain contrastive learning. While leveraging unlabeled image regions is essential, we introduce segmentation self-training with a novel instance-aware pseudo-label (IPL) selection strategy. Unlike existing methods that typically rely on pixel-wise pseudo-label filtering, the IPL semantically selects reliable and diverse pseudo-labels with the help of the detection task. Comprehensive validations and comparisons on challenging datasets demonstrate that our method outperforms existing UDA and WDA methods, significantly narrowing the performance gap with the supervised upper bound. Furthermore, under the UDA setting, our method also achieves substantial improvements over other UDA techniques.
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 →International journal of biometeorology
Recent trajectories of biometeorology in the Global South: a systematic review of biometeorology research in the International Journal of Biometeorology
Literature reviews have become increasingly prominent in scientific discourse; however, few have critically examined the geographic distribution of research, particularly in the field of biometeorology, where studies remain disproportionately concentrated in the Global North. Building on previous regional reviews and special issues, this paper critically examines geographic disparities in publication trends, authorship, and methodological approaches to aid in identifying challenges and opportunities for equitable and justified future research. Using a PRISMA-style framework, 591 empirical studies conducted across 147 Global South countries were identified from a total of 3,046 publications in the International Journal of Biometeorology (IJBM) between 2000 and 2024. The review reveals a steady increase in contributions from the Global South over time, but highlights persistent gaps in representation, especially in author affiliation and research leadership. Thematic and methodological analyses highlight the diversity and innovation present in Global South contexts, while also revealing structural barriers to equitable participation in the field. This work advocates for more inclusive research practices and enhanced institutional support to bridge the divide between the Global North and South, fostering a more representative and collaborative future for biometeorology.
12 May 2026
Read appraisal →Neuroscience
In silico neuronal morphology classification: A systematic review
Advances in connectomics and the characterization of neuronal diversity have been fundamental to understanding how the brain works. Defining a taxonomy is still challenging and requires complex computational methods. In this paper, we present a systematic review of the state-of-the-art methods for neuronal morphology classification, shedding light on trends and future directions. We address three core research questions: (1) What computational methods are used to classify neuronal morphologies? (2) What kind of representation do the classification models work on? (3) How are the classification models evaluated? We applied a structured protocol including peer-reviewed studies published since 2018. We collected 840 papers from five databases, published between 2018 and 2024. We assessed risk of bias using predefined quality criteria on study design, methodology, and reporting; 35 studies met these standards and were included to extract information and consolidate the results. The results show that Artificial Neural Networks are currently the main method for classifying neurons based on morphometric features (21 papers). Representations based on microscopy images for extracting morphometric features (30) are commonly used. Accuracy (29) is the predominant evaluation measure. Our research findings contribute to the body of knowledge by providing an up-to-date perspective on neuronal classification, which can guide future research in neuroscience and related fields.
11 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 →Developmental science
Making Machine Learning Accessible for Developmental Science: The Case of Automated Face Detection
The last decade has seen rapid advancements in machine learning, significantly transforming fields like cybersecurity and healthcare. Developmental science has been slower to adopt these technologies. Yet, machine learning holds immense potential to transform this field, enabling scalable and data-driven insights into developmental processes. Broader adoption is currently hindered by challenges in algorithm selection and technical implementation. We address these barriers by focusing on an area that has reached high sophistication from a machine learning perspective while also being of significant interest to developmental scientists: face detection. Face detection is crucial for analysing visual experiences through children's dynamic, first-person views. Automatising this process allows efficient handling of large egocentric datasets, enabling well-powered studies otherwise limited by labour-intensive manual annotation. Here, we systematically evaluated 13 state-of-the-art face detection algorithms (DeepFace library) using data from two increasingly common developmental methodologies involving children under 3 years of age: head-mounted eye-tracking in more structured settings (N = 20; n = 10 4-month-olds, n = 10 8-month-olds) and head-mounted cameras in naturalistic home environments (N = 10 18-29-month-olds). Benchmarking these algorithms against manual annotations revealed that YOLOv11Face (M) and RetinaFace consistently outperformed others in terms of precision and recall, exhibiting strong concordance with manual ratings, lower error, reduced systematic deviation and robust rank-order correlations with manual annotations. To facilitate broader adoption, we introduce an accessible face detection tool (TinyExplorer Detection App), promoting efficiency, scalability, and innovation in developmental science by widening access to machine learning.
5 May 2026
Read appraisal →Journal of clinical ultrasound : JCU
Illuminating Research Dynamics: Medical Ultrasound and Deep Learning
BACKGROUND: Deep learning (DL) has enabled advances in ultrasound imaging, but challenges like limited datasets and device variability hinder progress. This study provides the first bibliometric overview of DL research in medical ultrasound. METHOD: We retrieved related publications (2004-Apr 2025) on medical ultrasound and DL from the Web of Science Core Collection. Bibliometric tools (Bibliometrix, VOSviewer, CiteSpace) were used to quantify publication growth, country/institution contributions, and thematic evolution. RESULTS: A total of 3386 publications were included, with the annual output increasing steadily at an average growth rate of 31.02%. While China led all countries in publication volume (1355), the United Kingdom achieved the highest citation rate per article (55.5 citations). Fudan University was the most prolific institution (110), and key authors such as Luca Saba and Jasjit S. Suri formed the principal collaboration hubs. In terms of venues, Diagnostics published the most papers, whereas IEEE Transactions on Medical Imaging exerted the greatest citation impact (5646 citations; IF 8.9). Keyword analysis revealed core themes in classification, segmentation, and CNN-based diagnosis. From 2020 onward, ultrasound-specific topics (e.g., diagnosis, images) dominate; advanced architectures (e.g., U-Net) and emerging clinical concepts such as noninvasive assessment. CONCLUSION: This mapping highlights a rapidly growing, internationally collaborative field. Emerging trends include advanced artificial intelligence (AI) architectures (U-Net, attention-based networks, vision Transformers) and a shift toward surgical and therapeutic applications. These insights underscore the need for multicenter validation, standardized data protocols, and clinically interpretable AI models to accelerate safe translation into practice.
5 May 2026
Read appraisal →Nature reviews. Genetics
Interpretation, extrapolation and perturbation of single cells.
Single-cell analyses have transitioned from descriptive atlasing towards inferring causal effects and mechanistic relationships that capture cellular logic. Technological advances and the growing scale of observational and interventional datasets have fuelled the development of machine learning methods aimed at identifying such dependencies and extrapolating perturbation effects. Here, we review and connect these approaches according to their modelling concepts (including representation learning, causal inference, mechanistic discovery, disentanglement and population tracing), underlying assumptions and downstream tasks. We propose a unifying ontology to guide practitioners in selecting the most suitable methods for a given biological question, with detailed technical descriptions provided in an online resource . Finally, we identify promising computational directions and underexplored data properties that could pave the way for future developments.
4 May 2026
Read appraisal →The Proceedings of the Nutrition Society
Good practices and common pitfalls of machine learning in nutrition research
Machine learning is increasingly being utilised across various domains of nutrition research due to its ability to analyse complex data, especially as large datasets become more readily available. However, at times, this enthusiasm has led to the adoption of machine learning techniques prior to a proper understanding of how they should be applied, leading to non-robust study designs and results of questionable validity. To ensure that research standards do not suffer, key machine learning concepts must be understood by the research community. The aim of this review is to facilitate a better understanding of machine learning in research by outlining good practices and common pitfalls in each of the steps in the machine learning process. Key themes include the importance of generating high-quality data, employing robust validation techniques, quantifying the stability of results, accurately interpreting machine learning outputs, adequately describing methodologies, and ensuring transparency when reporting findings. Achieving this aim will facilitate the implementation of robust machine learning methodologies, which will reduce false findings and make research more reliable, as well as enable researchers to critically evaluate and better interpret the findings of others using machine learning in their work.
4 May 2026
Read appraisal →Scientific reports
scDecorr: feature decorrelation based representation learning enables self-supervised alignment of multiple single-cell experiments
Single-cell RNA sequencing (scRNA-seq) has revolutionized our understanding of cellular heterogeneity in complex biological systems. However, analyzing and integrating scRNA-seq data poses unique computational challenges due to sparsity, high variability, and technical batch effects. Here, we propose a novel framework called scDecorr for robust representation learning and data integration for scRNA-seq analysis. Our approach leverages the idea of feature decorrelation-based self-supervised learning (SSL) to obtain efficient low-dimensional representations of individual cells without relying on cell-type annotations. By maximizing similarity among distorted embeddings while decorrelating their components, scDecorr captures the biological signature while eliminating technical noise. Furthermore, scDecorr incorporates unsupervised domain adaptation to bridge the gap between batches with different distributions, enabling effective integration of scRNA-seq data from diverse sources. Our framework achieves domain-invariant representations by learning cell embeddings independently across domains and employing domain-specific batch normalization. We evaluate scDecorr on a variety of single-cell datasets and demonstrate its ability to integrate batches without losing the inherent biological variance, thereby facilitating optimal clustering. The representations generated by scDecorr also exhibit robustness in label transfer tasks, allowing for effective transfer of cell-type labels from reference to query datasets. Overall, scDecorr offers a powerful tool for efficient analysis and integration of large and complex scRNA-seq datasets, advancing our understanding of cellular processes and disease mechanisms. The code is available here https://github.com/hayatlab/scdecorr .
3 May 2026
Read appraisal →Turkish journal of ophthalmology
Large Language Models in Ophthalmology: A Bibliographic Analysis
This study evaluated the distribution of research on the use of large language models (LLMs) in ophthalmology through a bibliographic analysis of articles retrieved from PubMed through November 2024. Studies were categorized into four main areas of LLM application: clinical decision-making (further divided according to subspecialties), education, patient interactions, and miscellaneous applications. Descriptive statistics were used to analyze the distribution of studies by ophthalmic subspecialty, geographical region, journal quality, and author characteristics, including gender and scholarly impact (h-index and i10-index). The findings revealed that clinical decision-making was the most common application (43.7%), with the majority of studies in this subgroup focusing on the retina (39.5%). Geographically, most of the research originated from North America (48.3%), followed by Asia (29.9%) and Europe (20.7%). Most studies were published in high-impact journals (Q1 journals: 74.7%), particularly for those related to clinical decision-making in retina (80.0%), glaucoma (100%), and multiple subspecialties (87.5%). Gender disparities were evident across all author roles, with female authors accounting for only 29.9% of first authors, 25.3% of last authors, and 26.4% of corresponding authors. The results suggest a need for greater diversity in terms of gender and geographic representation in LLM research in ophthalmology to promote inclusive progress in the field. Bu çalışma, PubMed’de Kasım 2024’e kadar yayımlanmış göz hastalıkları alanında büyük dil modellerinin (BDM) kullanımı ile ilgili araştırmaların bibliyografik analizini amaçlamaktadır. Çalışmalar, klinik karar desteği (yandallara göre ayrılmış), eğitim, hasta etkileşimleri ve diğer uygulamalar olmak üzere BDM yöntemlerinin dört ana uygulama alanına göre kategorize edilmiştir. Tanımlayıcı istatistikler kullanılarak çalışmaların, oftalmoloji alt uzmanlıkları, coğrafi bölge, dergi kalitesi ve yazar özellikleri (cinsiyet ve bilimsel etki, h-endeksi ve i10-endeksi dahil) açısından dağılımı analiz edilmiştir. Bulgularımız, klinik karar desteğinin en yaygın uygulama olduğunu (%43,7) ve bu alt grup içerisinde çalışmaların çoğunun retina üzerine odaklandığını (%39,5) ortaya koymuştur. Coğrafi olarak, araştırmaların çoğu Kuzey Amerika’dan (%48,3) yayımlanmakta olup sırasıyla Asya (%29,9) ve Avrupa (%20,7) Kuzey Amerika’yı izlemektedir. Çoğu çalışma, özellikle de klinik karar desteği alt grubunda, retina (%80,0), glokom (%100) ve birden fazla alt uzmanlık dalını kapsayan çalışmalar (%87,5), yüksek etki faktörüne sahip dergilerde yayımlanmıştır (Q1 dergileri: %74,7). Cinsiyet dağılımındaki eşitsizlik, tüm yazar rollerinde belirgin olarak gözlemlenmiştir. Çalışmalarda kadın yazarlar, ilk yazarların yalnızca %29,9’unu, son yazarların %25,3’ünü ve sorumlu yazarların %26,4’ünü oluşturmuştur. Sonuçlar, göz hastalıkları alanındaki BDM araştırmalarında cinsiyet ve coğrafi temsilde daha fazla çeşitliliğe ihtiyaç duyulduğunu ve bu çeşitliliğin alandaki kapsayıcı ilerlemeyi teşvik edeceğini göstermektedir.
2 May 2026
Read appraisal →Journal of chemical information and modeling
TriPDCL: A Tri-Pathway Prototype-Driven Contrastive Learning Framework for Cross-Modality Single-Cell Integration
The rapid development of single-cell sequencing technologies has laid a solid foundation for multiomics data integration. A central objective in multiomics integration is to attenuate cross-omic discrepancies stemming from technical batch effects and inherent biological variability, thereby facilitating a unified latent representation that supports robust downstream analyses. However, the intrinsic high sparsity of single-cell data and pronounced intercellular heterogeneity pose greater challenges for joint analysis. We foreground cellular heterogeneity and present TriPDCL, a prototype-based contrastive learning framework, which not only achieves effective transfer of heterogeneity information through an iterative prototype-learning update mechanism but also, by leveraging learnable prototype centers, enables precise construction of reliable positive-negative sample pairs. It tackles two principal challenges, the alignment of cross-modal heterogeneity and the robust learning, in multiomics analysis. Subsequently, comparative assessments conducted on five data sets in contrast to seven representative methods illustrate the superiority of the proposed approach.
2 May 2026
Read appraisal →BMJ open
Do studies of interventions to improve laypeople's critical thinking about health choices assess potential harms? A systematic review.
OBJECTIVES: To make informed health choices, and avoid waste and unnecessary suffering, people need critical thinking skills. However, like health interventions, educational interventions can have adverse effects. In this systematic review, the objective was to assess the extent to which researchers have included potential adverse effects in studies of interventions intended to improve the critical thinking of laypeople about health choices. DESIGN: This study was a systematic review, in which we updated the search for an earlier systematic review of intended effects of relevant interventions. The earlier review did not address potential adverse effects. We did not update the analysis of intended effects. DATA SOURCES: We searched Cochrane Central Register of Controlled Trials (CENTRAL), Cumulative Index to Nursing and Allied Health Literature (CINAHL), Embase, Epistemonikos, Medical Literature Analysis and Retrieval System Online (MEDLINE), Education Resources Information Center (ERIC) and Web of Science up to March 2025. In addition to studies from the original review and updated search, we included any additional studies included in a similar, even earlier review. Our unit of analysis was study report (eg, journal article). ELIGIBILITY CRITERIA: We included all studies from the original review. We applied the same inclusion criteria to the results of the updated search: the study included a comparison, the population was laypeople and the intervention was intended to improve understanding of ≥1 key concept for informed health choices. DATA EXTRACTION AND SYNTHESIS: We extracted data about study design (randomised trial or other), participants (children, adolescents or adults), study setting (countries), main intervention (resources delivered to participants) and comparator (usual/no intervention or other). For the analysis, we extracted verbatim text describing any assessment of a potential adverse effect of the intervention. We conducted a narrative synthesis of the extracted data. RESULTS: We included 35 reports of quantitative studies (including multi-method and mixed-methods studies). Most often, the study was a randomised trial, the setting was a high-income country, the population included adults (including university students) and the intervention was school-based (including university). In one of the 35 reports, authors described assessing a potential adverse effect. CONCLUSION: To our knowledge, this is the first systematic review assessing the extent to which researchers have assessed adverse effects of any category of educational interventions. Our review shows that researchers generally have not assessed potential adverse effects of interventions to improve critical thinking about health choices. Researchers should pay more attention to such effects, while policymakers and educators making decisions about implementing relevant interventions should consider the lack of evidence. The findings of this study suggest a need for research that facilitates assessing potential adverse effects of interventions to improve critical thinking about health choices.
27 Apr 2026
Read appraisal →PloS one
Comparison of deep learning-based three-dimensional human pose estimation methods with motion capture for gesture research
Humans often make gestures while speaking. These gestures have been extensively researched in psychology and cognitive science, revealing their functions and roles in communication. Despite being produced in three-dimensional (3D) space, gestures and gesture space have primarily been measured and analyzed using two-dimensional planes. Optical motion capture (MoCap) is commonly used for 3D gesture measurement. However, MoCap is expensive, requires a large space, and the attachment of retroreflective markers can interfere with the natural generation of gestures by the speaker. Recent advances in deep learning-based human pose estimation (HPE) offer promising alternatives by enabling 3D keypoint estimation from standard video cameras. This study investigates the accuracy of four HPEs-two monocular and two stereo with triangulation methods-in estimating upper-body keypoints commonly used in natural scenarios for gesture research. Ten participants were recorded performing gesture-rich speech by both a MoCap system and three video cameras. We compared 13 keypoints, including the wrists, elbows, shoulders, fingers, and face, with MoCap data, using Euclidean distance as an error metric. Statistical analyses revealed that stereo methods significantly outperformed monocular methods across all keypoints. The most accurate method achieved an average error of 49.4 mm with respect to MoCap, suggesting sufficient accuracy for practical gesture analysis. Additionally, we visualized 3D gesture spaces and found a 75.4% overlap between HPE and MoCap at a voxel size of 50 mm, indicating high spatial agreement. Our findings demonstrate that stereo HPE methods are viable, cost-effective alternatives to MoCap in gesture-related applications. Our ultimate objective is to create a toolbox that is accurate and easy to use for measuring human gestures in 3D space by exploiting the recent advances in HPE as an alternative to MoCap. Such a toolbox will be suitable for non-experts in machine learning, and these results lay the foundation for it.
27 Apr 2026
Read appraisal →PloS one
The application of large language models in meteorology graduate research: current status, impact, and prospects
With the rapid development of generative artificial intelligence, large language models (LLMs) have gradually integrated into various fields, demonstrating significant potential, particularly in meteorological research. This study explores the current application, advantages, challenges, and future development trends of LLMs in the scientific work of meteorology graduate students at NUIST. Through surveys and case analysis, the study finds that LLMs are primarily applied in literature review, data processing, code development, and academic writing in meteorological research. The results show that LLMs significantly enhance research efficiency, particularly in code development and literature translation, saving considerable time for graduate students. However, challenges remain in areas such as the accuracy of professional knowledge, creative inspiration, and interdisciplinary integration. The study also reveals concerns over data security, academic integrity, and model limitations when using LLMs. Future applications of LLMs in meteorology need further optimization in terms of professional knowledge accuracy and data processing capabilities. This paper provides both theoretical support and practical guidance for the responsible integration of LLMs into meteorological research and education.
27 Apr 2026
Read appraisal →Behavior research methods
FaceTrack-AOI: An AI-driven tool for automated dynamic AOI placement and eye movement analysis in face perception studies
Face perception is fundamental to human social interaction, relying on the dynamic allocation of visual attention to facial features across natural variations in poses, expressions, lighting and context-critical factors for real-world face perception. While eye-tracking studies have advanced our understanding of this process, analyzing gaze data in naturalistic settings presents challenges due to the labor-intensive and often subjective process of manually defining areas of interest (AOI) in dynamic stimuli. We introduce FaceTrack-AOI, an open-source, artificial intelligence (AI)-driven tool that automates dynamic AOI construction using robust 68-point facial landmark detection. Featuring a user-friendly graphical interface, the tool allows researchers to define and export eye movement metrics without programming expertise. It supports both rectangular and polygon-based AOIs, with customizable buffers, blank zones, and hierarchical configurations, accommodating diverse theoretical and methodological needs. Validation on multi-ethnic naturalistic video frames, images from a standardized face dataset (Chinese Face and Body Dataset [CFBD]), and benchmarks against a published study revealed strong agreement with expert-defined manual AOIs, achieving a mean spatial overlap of 95.73%. A proof-of-concept validation confirmed high fixation capture: over 98.8% of dwell time was correctly directed toward key AOIs. In two experiments with infants (6-18 months) and adults viewing 28,800 dynamic video frames, and children (2-8 years), both neurotypical and with autism spectrum disorder (ASD), viewing 80 static faces, FaceTrack-AOI demonstrated enhanced sensitivity in detecting developmental and group differences in naturalistic face recognition. By integrating AI-powered tracking with an intuitive interface, FaceTrack-AOI streamlines gaze analysis for dynamic facial stimuli, offering researchers an efficient and scalable solution to advance face perception research in ecologically valid contexts.
26 Apr 2026
Read appraisal →Lab on a chip
Machine learning-driven single-cell phenotyping in size-controlled microenvironments via parallel deterministic droplet microfluidics
Understanding how individual cells respond to distinct physical microenvironments is critical for mechanobiology, cell therapy, and tissue engineering. Current single-cell encapsulation methods are often limited by Poisson loading and fixed droplet sizes, preventing parallel generation of multiple, size-specific microenvironments and constraining high-resolution phenotypic analyses. Here, we present a droplet microfluidic platform that enables deterministic single-cell encapsulation within microgels of multiple sizes from a single precursor stream, achieved through parallelized flow-focusing combined with cell-selective gelation. This system produces distinct microgel size regimes simultaneously, minimizing empty compartments and enabling direct, side-by-side comparisons of cellular behavior under controlled yet variable confinement. Using machine learning to analyze 3D morphological and cytoskeletal features, we reveal heterogeneous, size-dependent phenotypic responses and demonstrate that cellular phenotypes alone can predict microgel confinement across time. Together, these results establish a data-driven framework for mapping single-cell responses across engineered microenvironments and provide a scalable platform for predictive studies of mechanosensitive behavior in heterogeneous niches.
24 Apr 2026
Read appraisal →International journal of epidemiology
Structured tools for assessing quality and risk of bias in Mendelian randomization studies: an updated systematic review
BACKGROUND: The growing use of Mendelian randomization (MR) has heightened the need for rigorous quality and bias assessment tools. A previous systematic review included studies published up to July 2021 identified 14 structured instruments for conducting, evaluating, and reporting MR studies. However, methodological developments have accelerated in the years since. METHODS: We updated the previous systematic review to include tools published between July 2021 and January 2025, applying the same search strategy and eligibility criteria. Two reviewers independently screened articles, extracted data, and mapped tool content to bias domains. RESULTS: We identified 15 additional articles, bringing the total to 29 tools. Of these 29, 19 provided structured evaluation tools. Twelve of the 19 evaluation tools were newly added in the present review, which addressed broader methodological domains beyond core instrumental variable assumptions, including genetic instrument selection, population stratification, sensitivity analyses, and dataset considerations. However, substantial variation in bias domains, structure, and scoring methods across tools persists. Key gaps remain in the assessment of linkage disequilibrium, missing data, and dynastic effects. CONCLUSIONS: While the number of structured tools has increased in recent years, the lack of standardization across tools makes it difficult to compare across systematic reviews of MR studies. Developing more complete and standardized evaluation frameworks and properly testing these tools in practice are important next steps to improve the overall quality of MR research.
20 Apr 2026
Read appraisal →Medicine
Publication trends of dexmedetomidine and perioperative cognitive dysfunction: A bibliometric analysis from 1995 to 2024
INTRODUCTION: Dexmedetomidine (Dex) has shown potential in mitigating perioperative cognitive dysfunction (POCD). This study aimed to systematically evaluate the research hotspots and development trends in Dex and POCD using bibliometric analysis. METHODS: A systematic search was conducted on the Web of Science Core Collection database to identify publications related to Dex and POCD from 1995 to 2024. Bibliometric analysis was performed using VOSviewer (V 1.6.20), CiteSpace (V 6.3.R1), and R V4.3.3. RESULTS: A total of 585 articles on Dex and POCD between 1995 and 2024 were analyzed. These articles involved 1688 authors from 1857 institutions across 140 countries or regions, published in 240 journals, and citing 12,881 sources. The annual growth rate of publications in this field was 16.21%, with a peak of 91 articles in 2021. China led in the number of articles (n = 379), multiple country publications (MCP = 29), and total citations = 4989). The USA ranked second with 74 articles, 18 MCPs, and 2608 citations, followed by South Korea with 29 articles and 533 citations. Harvard University published the most articles (n = 51). Wang Dong-Xin was the most influential author, with an h-index of 7. Anesthesiology led the high-impact journals with an H-index of 11. The keyword analysis identified "neuroinflammation," "risk," "injury," and "efficacy" as the latest research themes. CONCLUSION: This bibliometric analysis provides a comprehensive overview of the research trends regarding Dex and POCD. Future research should continue to explore the mechanisms, efficacy, and safety of Dex in preventing POCD.
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 →African journal of reproductive health
Mapping research fronts of reproductive medicine using the NHANES database: A bibliometric and thematic analysis
Reproductive health, infertility, and perinatal diseases significantly challenge human health development and continuity, requiring multidisciplinary approaches for enhanced academic understanding and practical solutions. This study used bibliometric analysis to examine related articles in the web of science (WoS) database, employing tools including bibliometrix and VOSviewer for systematic literature analysis. To capture the broader context of high-impact research, we first analyzed 1,026 highly-cited and hot papers in reproductive medicine, which revealed evolving thematic trends and global collaboration patterns. Then we examined 232 literatures using NHANES datasets for reproductive health research to identify research gaps and opportunities for further research in this field. The analysis revealed key trends, areas of concentration, and potential knowledge gaps in the literature, while the thematic analysis of reproductive health studies utilizing the NHANES database further underscored the wealth of data amenable to multidisciplinary investigation. Future research directions to explore new dimensions of reproductive health, improving health policies and clinical practices in this field using NHANES database are also suggested. The study concludes that multidisciplinary research using comprehensive datasets like NHANES is crucial for advancing understanding of reproductive health, infertility, and perinatal outcomes. La santé reproductive, l'infertilité et les maladies périnatales constituent un défi majeur pour le développement et la continuité de la santé humaine, nécessitant des approches multidisciplinaires afin d'améliorer la compréhension académique et de proposer des solutions pratiques. Cette étude a utilisé une analyse bibliométrique pour examiner les articles pertinents dans la base de données WoS, en employant des outils tels que Bibliometrix et VOSviewer pour une analyse systématique de la littérature. Afin de saisir le contexte plus large de la recherche à fort impact, nous avons d'abord analysé 1 026 articles hautement cités et à tendance chaude en médecine reproductive, ce qui a révélé les tendances thématiques évolutives et les modèles de collaboration internationale. Nous avons ensuite examiné 232 publications utilisant les jeux de données NHANES pour la recherche sur la santé reproductive, afin d'identifier les lacunes de recherche et les opportunités pour de futures études dans ce domaine. L'analyse a mis en évidence les tendances clés, les domaines de concentration et les lacunes de connaissances potentielles dans la littérature, tandis que l'analyse thématique des études sur la santé reproductive utilisant les jeux de données NHANES a en outre souligné la richesse des données susceptibles d'être étudiées de manière multidisciplinaire. Les orientations futures de la recherche visant à explorer de nouvelles dimensions de la santé reproductive, à améliorer les politiques sanitaires et les pratiques cliniques dans ce domaine à l'aide des jeux de données NHANES sont également proposées. L'étude conclut que la recherche multidisciplinaire utilisant des jeux de données complets tels que NHANES est cruciale pour faire progresser la compréhension de la santé reproductive, de l'infertilité et des issues périnatales.
17 Apr 2026
Read appraisal →Journal of chemical information and modeling
zERExtractor: An Automated Platform for Enzyme-Catalyzed Reaction Data Extraction from Scientific Literature
The rapid expansion of enzyme reaction literature has created a major bottleneck in database curation, leaving vast amounts of enzyme-substrate-condition relationships unstructured and inaccessible for DL-driven modeling. How to fully utilize the enzymatic reaction data has been an important task for future accurate enzyme activity prediction models. Current deep learning (DL)-based data extraction models heavily rely on large language models (LLMs) without a fidelity check and the ability to continuously evolve. To address these issues, we developed zERExtractor (Zelixir's Enzyme Reaction Data Extractor), an accuracy-oriented and extensible platform for extracting enzyme-catalyzed reaction data from scientific publications. This system offers a unified multimodal information extraction framework (covering molecular reaction diagrams, tables, and texts) to integrate enzymatic reaction descriptors into structured storage. We employ fine-tuned large LLMs together with DL in a human-in-the-loop pipeline that evolves through data fidelity validation by experts and active learning. Also, zERExtractor achieves 89.9% accuracy in table recognition and over 98% accuracy in molecular image recognition on synthetic data sets, outperforming the strongest baseline by more than 2% and consistently maintaining above 95% on realistic benchmarks. zERExtractor bridges the data gap in enzyme reaction data with a scalable framework for accurate multimodal extraction, advancing DL-driven enzyme modeling and enabling future applications in computational enzymology and biotechnology. The platform is publicly accessible online at https://zpaper.zelixir.com/.
14 Apr 2026
Read appraisal →Journal of chemical information and modeling
ncProFormer: A CNN-enhanced Transformer for ncRNA Coding-Potential Prediction
Recent studies have revealed that some noncoding RNAs (ncRNAs) bear translational potential, and their encoded micropeptides have essential functions in multiple biological processes. However, accurate identification of coding-capable ncRNAs remains challenging due to weak translation signals, low conservation, and heterogeneous data distributions. Herein, we propose ncProFormer, a deep learning framework tailored for ncRNA coding-potential prediction. ncProFormer integrates the nucleic-acid language model GENA-LM to obtain contextual sequence embeddings, adopts an all-token representation strategy, and employs a convolutional neural network (CNN)-enhanced transformer encoder to jointly capture local nucleotide patterns and long-range dependencies. ncProFormer consistently outperformed the existing methods across the in-house human data set, the external validation data set, the public CPPred benchmark data set. More importantly, this study presents the first cross-species evaluation in ncRNA coding-potential prediction. Without retraining, ncProFormer maintained its strong predictive performance on mouse and rat data sets, showing that the learned biological representations are transferable and it is robust under the distributional shift and cross-species conditions. Collectively, these findings establish ncProFormer as an effective and generalizable framework for uncovering the coding potential of ncRNAs, thus offering a promising computational tool for characterizing ncRNA functions across diverse transcriptomic contexts.
14 Apr 2026
Read appraisal →Medicine
Bibliometric and visual analysis of artificial intelligence-related research in osteoarthritis: Trends and frontiers (2008-2025)
Osteoarthritis (OA) causes joint pain, functional impairment, and significantly impacts patients' quality of life. Recent advancements in artificial intelligence (AI) technology have provided new opportunities for OA diagnosis, treatment, and management. This study aims to understand the knowledge structure and research frontiers of "AI and OA" using bibliometric methods. Literature on "AI and OA" from January 1, 2008, to March 31, 2025, was retrieved from the Web of Science Core Collection database. VOSviewer, CiteSpace, and the R package "bibliometrix" were used for analysis. The study included 659 articles from 59 countries. Since 2019, the number of scientific outputs related to OA and AI has increased annually. The United States, China, and the United Kingdom are the main contributing countries, with Harvard University, Seoul National University, and the University of California system as leading institutions. The "JOURNAL OF ARTHROPLASTY" and "OSTEOARTHRITIS AND CARTILAGE" are the most contributing journals. Among the 2404 authors, GUERMAZI A and ROEMER FW are the most influential. Thematic trend analysis shows that AI and machine learning are driving new trends in OA research, especially in diagnosis, treatment planning, and patient monitoring. This paper summarizes the bibliometric trends and development in "AI and OA" studies over 18 years, identifying the latest research frontiers and hot topics. It provides valuable references for related research and highlights the importance of interdisciplinary collaboration as research topics diversify.
12 Apr 2026
Read appraisal →PloS one
Comparative analysis of text readability and writing styles in AI-generated vs. Human-written academic abstracts
Research article abstracts are vital in scientific publications for readers to assess a study's significance. The increasing use of AI tools, such as Kimi, ChatGPT and DeepSeek, to generate abstracts raises concerns about their readability and writing styles compared to human-written ones. The study aims to compare the differences in text readability and writing styles between human-written against AI-generated abstracts. A total of 150 abstracts of high-impact journal articles in the field of linguistics and computer science, 75 from each discipline, and another 150 AI-generated abstracts from the same corpus of articles served as the source texts for analysis. The Readability Scoring System, a computational tool, yielded readability and writing style metrics, while expert evaluation was performed to assess the quality of AI-generated academic abstracts. The quantitative data generated were analysed using SPSS 27 with non-parametric statistical methods. Key findings revealed: (1) AI-generated abstracts exhibited significantly lower readability across eight metrics, indicating greater complexity and lower readability; (2) Discipline-specific analysis showed five differing metrics in linguistics and eight in computer science; (3) Interdisciplinary comparisons revealed non-significant differences across nine readability metrics, highlighting AI's potential to mimic natural writing. However, it still faces challenges in generating lexically diverse content. These results underscored the current limitations of AI in generating readable and human-like abstracts, especially in technical fields.
10 Apr 2026
Read appraisal →Genome research
spRefine denoises and imputes spatial transcriptomic data with a reference-free framework powered by genomic language model
The analysis of spatial transcriptomic data is hindered by high noise levels and missing gene measurements, challenges that are further compounded by the higher cost of spatial data compared to traditional single-cell data. To overcome this challenge, we introduce spRefine, a deep learning framework that leverages genomic language models to jointly denoise and impute spatial transcriptomic data. Our results demonstrate that spRefine yields more robust cell- and spot-level representations after denoising and imputation, substantially improving data integration. In addition, spRefine serves as a strong framework for model pretraining and the discovery of novel biological signals, as highlighted by multiple downstream applications across data sets of varying scales. Notably, spRefine enhances the accuracy of spatial aging clock estimations and uncovers new aging-related relationships associated with key biological processes, such as neuronal function loss, which offers new insights for analyzing aging effect with spatial transcriptomics.
8 Apr 2026
Read appraisal →Regional anesthesia and pain medicine
Human versus artificial intelligence: evaluating ChatGPT's performance in conducting published systematic reviews with meta-analysis in chronic pain research
INTRODUCTION: Artificial intelligence (AI), particularly large-language models like Chat Generative Pre-Trained Transformer (ChatGPT), has demonstrated potential in streamlining research methodologies. Systematic reviews and meta-analyses, often considered the pinnacle of evidence-based medicine, are inherently time-intensive and demand meticulous planning, rigorous data extraction, thorough analysis, and careful synthesis. Despite promising applications of AI, its utility in conducting systematic reviews with meta-analysis remains unclear. This study evaluated ChatGPT's accuracy in conducting key tasks of a systematic review with meta-analysis. METHODS: This validation study used data from a published meta-analysis on emotional functioning after spinal cord stimulation. ChatGPT-4o performed title/abstract screening, full-text study selection, and data pooling for this systematic review with meta-analysis. Comparisons were made against human-executed steps, which were considered the gold standard. Outcomes of interest included accuracy, sensitivity, specificity, positive predictive value, and negative predictive value for screening and full-text review tasks. We also assessed for discrepancies in pooled effect estimates and forest plot generation. RESULTS: For title and abstract screening, ChatGPT achieved an accuracy of 70.4%, sensitivity of 54.9%, and specificity of 80.1%. In the full-text screening phase, accuracy was 68.4%, sensitivity 75.6%, and specificity 66.8%. ChatGPT successfully pooled data for five forest plots, achieving 100% accuracy in calculating pooled mean differences, 95% CIs, and heterogeneity estimates (I2 score and tau-squared values) for most outcomes, with minor discrepancies in tau-squared values (range 0.01-0.05). Forest plots showed no significant discrepancies. CONCLUSION: ChatGPT demonstrates modest to moderate accuracy in screening and study selection tasks, but performs well in data pooling and meta-analytic calculations. These findings underscore the potential of AI to augment systematic review methodologies, while also emphasizing the need for human oversight to ensure accuracy and integrity in research workflows.
4 Apr 2026
Read appraisal →Aesthetic Plast Surg
The Rise of Intelligent Plastic Surgery: A 10-Year Bibliometric Journey Through AI Applications, Challenges, and Transformative Potential
BACKGROUND: Driven by advancements in deep learning, surgical robots, and predictive modeling technologies, the integration of artificial intelligence (AI) and plastic surgery has expanded rapidly. Although AI shows the potential to enhance precision and efficiency, its clinical integration faces challenges, including ethical concerns and interdisciplinary complexity, which require a systematic analysis of research trends. METHODS: The CiteSpace and VOSviewer software were used to conduct a quantitative analysis of 235 documents in the core collection of Web of Science from 2016 to 2024. Co-citation networks, keyword co-occurrence, burst detection, and cluster analysis were employed to map the research trajectories. The inclusion criteria gave priority to studies that explicitly incorporated artificial intelligence into surgical designs or outcomes. The contributions of countries, institutions, and authors were evaluated through centrality indicators. RESULT: Publications related to artificial intelligence have grown exponentially, with the USA, Germany, and Canada leading research output. Harvard and Stanford Universities dominate in terms of institutional contributions, but cross-institutional collaboration remains limited. The keyword cluster highlights the innovations of artificial intelligence in breast reconstruction, facial analysis, and automated grading systems. Burst terms such as "deep learning," "risk assessment," and "attractiveness" underscore AI's role in optimizing surgical outcomes, but they also expose biases against Western-centric beauty standards. Ethical concerns, dataset diversity gaps, and overreliance on AI-driven decisions have become key obstacles. CONCLUSION: The integration of artificial intelligence in plastic surgery goes beyond the utility based on tools and into data-informed surgical engineering. The persistent gap in collaboration and dataset diversity highlights the need for global, interdisciplinary efforts to address technical and ethical challenges while advancing AI's clinical utility. Future research must prioritize transparency, inclusivity, and collaborative innovation to realize AI's transformative potential while mitigating risks. LEVEL OF EVIDENCE IV: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .
1 Mar 2026
Read appraisal →Int J Surg
Artificial Intelligence in melanoma research: a bibliometric analysis
BACKGROUND: As one of the most lethal skin cancers, melanoma has encountered many obstacles in diagnosis and therapy. Artificial intelligence (AI) can help improve early diagnosis, prognosis, and treatment of melanoma. However, there is a lack of detailed and accurate bibliometric analysis of the field. METHODS: All publications were extracted from Web of Science Core Collection based on AI and melanoma terms. Bibliometric analysis was conducted on 1476 articles/reviews by using VOSviewer, CiteSpace, and bibliometrix for co-authorship, citation, keyword, and journal analysis. RESULTS: There was a sudden increase in publications each year after 2017 and reached 260 in 2024. The United States (337 publications) and China (292 publications) ranked top in publication productivity. Germany was the top institution and author country. IEEE Access (57 publications) and Diagnostics (54 publications) were core journals. The two most prominent research hotspots were AI-assisted diagnosis and AI-integrated immunotherapy according to the keyword analysis. CONCLUSION: AI in melanoma research has exploded since 2017. It is recommended that future research focus on various datasets, explainable AI, and cross-disciplinary cooperation to promote the transformation of achievements into clinical practice.
1 Feb 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 →