Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 25 appraisals
Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophysiology
Handwritten signatures in neurological and psychiatric disorders: a systematic review with clinical and forensic implications.
BACKGROUND: Signature production represents a complex motor and cognitive behavior, highly automatized and repeated countless times throughout life. It ranks among the most durable graphic forms, often preserved even when writing processes are impaired. Studies examining signature characteristics in clinical cohorts are limited. The available evidence raises questions about its potential as a behavioral marker of neurological and psychiatric disorders. METHODS: We conducted a PRISMA-compliant systematic review up to 4th December 2025 to summarize evidence on how neurological and psychiatric conditions may affect signature features. Eligible articles on this topic were searched in Scopus, Web of Science, and PubMed databases. Sixteen papers were included, and their results were systematically described and qualitatively analyzed. RESULTS: The available literature suggests that signature production appears relatively independent of global cognitive decline, although it is sensitive to specific pathological conditions (e.g., micrographia in Parkinson's disease, increased tortuosity in cerebral small vessel disease, and alterations observed in mania, delirium, obsessive-compulsive disorder, and major depressive disorder). Alterations commonly affect kinematic parameters, as well as visuospatial and constructional features. However, no disorder-specific signature pattern emerges from current evidence, largely due to the methodological heterogeneity across studies. CONCLUSIONS: The handwritten signature holds promise as a behavioral marker of brain disorders, serving as a potentially useful, complementary - rather than a stand-alone - indicator of neurological and psychiatric conditions, which could be potentially useful in clinical settings. The implications for forensic applications remain largely theoretical and have not yet been standardized.
30 July 2026
Read appraisal →Neuropsychologia
Number of types of synaesthesia can be predicted by structural and functional neuroimaging data.
It has been suggested that synaesthetes have a distinct neurocognitive profile with a broad variety of cognitive and behavioural differences. Recent studies have shown that people with more types (relative to fewer types) of synaesthesia are easier to classify using machine learning of questionnaires and cognitive test data. This suggests a spectrum within synaesthesia, despite synaesthesia itself being typically defined in categorical terms. This study uses the same basic approach applied to 13 brain-based biomarkers. These have previously been shown to distinguish synaesthetes from controls, but it is not known whether they explain heterogeneity amongst synaesthetes. Using machine learning methods (elastic net regression), we were able to find several biomarkers that predict above chance the number of types of synaesthesia. These include both functional MRI (the extent to which brain regions act as hubs) and structural MRI (e.g., intracortical myelination) measures. This is the first project that explores whether it's possible to predict the breadth of synaesthesia from brain-based measures.
29 July 2026
Read appraisal →Science advances
Curiosity-driven development of action and language in robots through self-exploration
Infants acquire language with generalization from minimal experience, whereas large language models require billions of training tokens. What underlies efficient development in humans? We investigated this problem through experiments wherein robotic agents learn to perform actions associated with imperative sentences (e.g., push red cube) via curiosity-driven self-exploration. Our approach amortizes active inference using Q-learning, enabling intrinsically motivated developmental learning. The simulations reveal key findings corresponding to observations in developmental psychology. (i) Generalization improves markedly as the scale of compositional elements increases. (ii) Curiosity-driven exploration enables faster learning. (iii) Rote pairing of sentences and actions precedes compositional generalization. (iv) Exception handling induces U-shaped developmental performance, a pattern like representational redescription in child language learning. These results suggest that curiosity-driven active inference accounts for how intrinsically motivated sensorimotor-linguistic learning supports scalable compositional generalization and exception handling in humans and artificial agents.
25 July 2026
Read appraisal →Behavior research methods
The English spaced compound word database
Compound words combine existing words to form a new concept. Spaced compound words represent one form of compound words in which the orthographic space is preserved between the two constituent words (e.g., coffee table). While recent research has mostly focused on unspaced compound words, spaced compound words offer important insight into the cognitive processes that support lexical retrieval and comprehension. Critically, whether readers process spaced compound words as single lexical units has important implications for how these items are stored and accessed in memory. To support future research, we developed the English Spaced Compound Word Database, wherein we collected familiarity, age-of-acquisition (AoA), and semantic transparency (ST) ratings for 1,162 spaced compound words. We conducted correlational analyses to compare these ratings across words, as well as across previously collected data on compound words' constituents. These analyses revealed a strong association overall between familiarity ratings and AoA, as well as familiarity and ST. When examining the relationship between ratings and the characteristics of a compound word's constituents, we found a strong relationship between frequency, AoA, and structural simplicity (i.e., number of characters, syllables, morphemes, and lexical similarity) within all three rating tasks. These findings provide a basis for testing future theoretical models of lexical processing and reading.
12 July 2026
Read appraisal →Philosophical transactions of the Royal Society of London. Series B, Biological sciences
The hippocampus as a generative model
A generative model can be defined as a model of the latent causes of sensory input that can be used to generate new data samples. By examining empirical evidence and computational theory, we propose that the hippocampus can be characterized as a generative model. The hippocampus is a brain region important for memory. Recordings of neural activity from the hippocampus have led to the view that the hippocampus represents a cognitive map by abstracting a low-dimensional representation of the external world. We extend this view to suggest the hippocampus represents the latent, unobserved causes of sensory data by virtue of the position of the hippocampus within the deep cortical hierarchy. These representations of unobserved latent causes endow the hippocampus with capacity to generate new data samples that allow exploration of future hypotheticals and provide an internally generated training signal back to the generative model. We explore how perturbations to the hippocampal generative model may explain core symptoms of neuropsychiatric disorders such as those observed in psychosis. Together, this perspective provides a unified account of hippocampal function that explains how computations performed by the hippocampus support higher-order cognition and adaptive behaviour. This article is part of the theme issue 'The role of hippocampal predictions in cognition: bridging perception and memory'.
11 July 2026
Read appraisal →PLoS computational biology
Predictive coding explains asymmetric connectivity in the brain: A neural network study
Seminal frameworks of predictive coding propose a hierarchy of generative modules, each attempting to infer the neural representation of the module one level below; the predictions are carried by top-down feedback projections, while the predictive error is propagated by reciprocal forward pathways. Such symmetric feedback connections support visual processing of noisy stimuli in computational models. However, neurophysiological studies have yielded evidence of asymmetric cortical feedback connections. We investigated the contribution of neural feedback in visual processing for computing grasp parameters, by utilizing convolutional neural network models that had been augmented with predictive feedback and were trained to compute grasp positions for real-world objects. After establishing an ameliorative effect of symmetric feedback on grasp detection performance when evaluated on noisy stimuli, we characterized the performance effects of asymmetric feedback, similar to that observed in the cortex. Specifically, we tested model variants extended with short-, medium-, long- and longer-range feedback connections (i) originating at the same source layer or (ii) terminating at the same target layer. We found that the performance-enhancing effect of predictive coding under adverse conditions was optimal for medium-range asymmetric feedback. Moreover, this effect was most prominent when medium-range feedback originated at a level of representational abstraction that was proximal to the input layer, in contrast to more distal layers. To conclude, our simulations show that introducing biologically realistic asymmetric predictive feedback improves model robustness to noisy visual stimuli in a neural network model optimized for grasp detection.
8 July 2026
Read appraisal →Behavior research methods
A validity-guided workflow for robust large language model research in psychology.
Large language models (LLMs) are rapidly being integrated into psychological and behavioral research as research tools, evaluation targets, human simulators, and cognitive models. Yet recent evidence reveals severe measurement unreliability: personality assessments degenerate under factor analysis, moral preferences reverse with punctuation changes, and theory-of-mind accuracy varies widely with trivial rephrasing. These "measurement phantoms"-statistical artifacts masquerading as psychological phenomena-threaten the validity of a growing body of research. Guided by the dual-validity framework that integrates psychometrics with causal inference, this article presents a six-stage workflow that scales validity requirements to research ambition-using LLMs to code text requires basic reliability and accuracy, whereas claims about psychological properties demand comprehensive construct validation. Researchers must (1) explicitly define their research goal and corresponding validity requirements, (2) develop and validate computational instruments through psychometric testing, (3) design experiments that control for computational confounds, (4) execute protocols transparently, (5) analyze data with methods appropriate for nonindependent observations, and (6) report findings within boundaries and use results to refine theory. The workflow is illustrated through an example of model evaluation-"LLM selfhood"-showing how systematic validation can distinguish genuine computational phenomena from measurement artifacts. By establishing validated computational instruments and transparent practices, this workflow provides a path toward building a robust empirical foundation for AI psychology research.
3 July 2026
Read appraisal →The Behavioral and brain sciences
Large language models have learned to use language.
Acknowledging that large language models have learned to use language can open doors to breakthrough language science. Achieving these breakthroughs may require abandoning some long-held ideas about how language knowledge is evaluated and reckoning with the difficult fact that we have entered a post-Turing test era.
2 July 2026
Read appraisal →The Behavioral and brain sciences
Large language models are not about natural language.
Large Language Models are useless for linguistics, as they are probabilistic models that require a vast amount of data to analyze externalized strings of words. In contrast, human language is underpinned by a mind-internal computational system that recursively generates hierarchical thought structures. The language system grows with minimal external input and can readily distinguish between real language and impossible languages.
2 July 2026
Read appraisal →eLife
Disinformation elicits learning biases.
In open societies, disinformation is often considered a threat to the very fabric of democracy. However, we know little about how disinformation exerts its impact, especially its influence on individual learning processes. Guided by the notion that disinformation exerts its pernicious effects by capitalizing on learning biases, we ask which aspects of learning from potential disinformation align with ideal 'Bayesian' principles, and which exhibit biases deviating from these standards. To this end, we harnessed a reinforcement learning framework, offering computationally tractable models capable of estimating latent aspects of a learning process as well as identifying biases in learning. In two experiments, participants completed a two-armed bandit task, where they repeatedly chose between two lotteries and received outcome-feedback from sources of varying credibility, who occasionally disseminated disinformation by lying about true choice outcome (e.g., reporting non-reward when a reward was truly earned or vice versa). Computational modelling indicated that learning increased in tandem with source credibility, consistent with ideal-Bayesian principles. However, we also observed striking biases reflecting divergence from idealized Bayesian learning patterns. Notably, in one experiment individuals learned from sources that should have been ignored, as these were known to be fully unreliable. Additionally, the presence of disinformation elicited exaggerated learning from trustworthy information (akin to jumping to conclusions) and exacerbated a normalized measure of 'positivity bias' whereby individuals self-servingly boost their learning from positive, relative to negative, choice feedback. Thus, in the face of disinformation we identify specific cognitive mechanisms underlying learning biases, with potential implications for societal strategies aimed at mitigating its harmful impacts.
28 June 2026
Read appraisal →Psychological medicine
Sorting the mind: cognitive enhancement through transcutaneous auricular vagus nerve stimulation: a systematic review and meta-analysis
Transcutaneous auricular vagus nerve stimulation (taVNS) is a noninvasive technique engaging vagal afferents that may enhance cognition, but results vary across domains and samples. Following PRISMA, seven databases (inception-October 2025) plus registries and gray literature were searched. Random-effects meta-analyses (REML; Hedges' g) were complemented by Bayesian hierarchical models and sensitivity analyses. Fifty-three studies were included; 30 contributed quantitative data (>1,500 participants). taVNS was associated with improved cognitive performance overall (g = 0.41, 95% CI: 0.30-0.53; I2 = 51.4%). Effects were moderate for executive functions (g = 0.46, 95% CI: 0.27-0.65; I2 = 9.5%) and cognitive flexibility/learning (g = 0.53, 95% CI: 0.32-0.75; I2 = 52.9%), and small for working memory/attention (g = 0.19, 95% CI: 0.04-0.33; I2 = 14.9%). Social cognition/emotion regulation showed larger but imprecise effects (k = 3; g = 0.80, 95% CI: 0.07-1.52; I2 = 82.1%). Clinical samples benefited similarly (k = 7; g = 0.55, 95% CI: 0.31-0.79; I2 = 29.5%), with no difference from healthy cohorts (β = -0.001, p = .994). High-intensity protocols (>1.0 mA) yielded larger effects; mode, duration, and site were not moderators. Bayesian models supported effects (P [μ > 0] ≥ 0.93). taVNS is associated with statistically significant improvements in cognitive performance, strongest for executive control and adaptive learning. We propose a Vagal Neurocognitive Integration Model linking LC-NE arousal modulation to prefrontal control. Future diagnosis-specific, adequately powered trials with multimodal neuroimaging should refine mechanisms and dose-response.
25 June 2026
Read appraisal →Behavioural brain research
Sex differences in the use of spatial strategies in humans: An updated systematic review of the last decade.
Although sex differences in spatial cognition have been widely studied, the extent to which men and women differ in their use of egocentric and allocentric spatial strategies remains unclear. This systematic review summarizes findings from 21 studies published between 2014 and 2024 that examined sex differences in spatial strategy use throughout development, adulthood, and aging. Studies were identified through searches of Scopus, Web of Science, PubMed, and PsycINFO, in accordance with PRISMA guidelines. Eligible articles examined egocentric and allocentric spatial strategies in humans and reported results by sex. Results demonstrated male advantages in allocentric strategy, as observed in thirteen studies involving children, young adults, and older adults. Men also outperformed women in several navigation outcomes, including target retrieval, route efficiency, and directional accuracy. Evidence for a female advantage in egocentric strategies was limited and inconsistent. Seven studies reported no sex differences, particularly those using small-scale, familiar, or less immersive tasks. This suggests that methodological factors and task demands strongly influence the observed effects. Age also influenced outcomes. Male advantages in allocentric strategies emerged in early childhood. Older women showed relatively preserved egocentric abilities, but greater decline in allocentric tasks. Men showed stronger hippocampal-parietal connectivity supporting allocentric processing, while women exhibited enhanced fronto-temporal connectivity. Hormonal studies highlight the context-dependent effects of testosterone and estrogen. These findings indicate that sex differences in spatial strategies are highly sensitive to task complexity, environmental familiarity, cognitive load, and individual factors. These findings emphasize the necessity of integrative models that consider developmental, neurobiological, and sociocultural contributors.
25 June 2026
Read appraisal →Progress in neuro-psychopharmacology & biological psychiatry
Neural correlates of late talking: A systematic review of electrophysiological and neuroimaging studies.
BACKGROUND: Late talking (LT), a delay in expressive vocabulary in early childhood, affects a significant minority of toddlers and can be a precursor to persistent language and literacy impairments. While behavioral profiles are well-documented, a systematic synthesis of its underlying neurobiology is lacking. This systematic review aimed to identify, evaluate, and synthesize evidence from electrophysiological (EEG) and neuroimaging (MRI) studies on the neural correlates of late talking in children. METHODS: We conducted a systematic search of Scopus, Web of Science, and PubMed from inception to September 2025, following PRISMA guidelines. Studies investigating children formally identified as late talkers without comorbid neurological or global developmental conditions, using EEG/ERP or MRI modalities, were included. Risk of bias was assessed using the Joanna Briggs Institute (JBI) critical appraisal tool. RESULTS: Sixteen studies (8 EEG, 8 neuroimaging) comprising 231 LT children in EEG studies and 315 in neuroimaging studies were included. EEG studies revealed a neurodevelopmental cascade, beginning with atypical oscillatory activity (increased frontal gamma power) and deficits in auditory discrimination (reduced Mismatch Negativity), progressing to higher-order impairments in phonological (absent PMN) and lexical-semantic (altered N400) processing. Neuroimaging studies consistently identified structural and functional anomalies within a distributed perisylvian network, including reduced gray matter in left temporal regions, atypical right-hemispheric lateralization, reduced activation in cortical-subcortical circuits, and altered white matter connectivity. These neural markers demonstrated significant predictive value for later language outcomes. CONCLUSION: Late talking is a clear neurodevelopmental disorder with a cascading pattern of neural dysfunction. The identified neural signatures are not just findings but have strong clinical potential as biomarkers for early identification and prognosis. This evidence advocates for a shift to a precision medicine approach, opening the door to novel, neurobiologically-informed interventions like neuromodulation that target the underlying brain circuits.
20 June 2026
Read appraisal →Brain research bulletin
Neural substrates of apathy in Parkinson's disease: A systematic review of structural and functional neuroimaging studies
BACKGROUND: Apathy is a frequent and disabling non-motor symptom in Parkinson's disease (PD), characterized by diminished motivation, reduced goal-directed behavior, and emotional indifference. Growing evidence suggests that apathy in PD represents a distinct neuropsychiatric syndrome that may coexist with depression and global cognitive impairment, while also showing partially dissociable clinical and neural correlates. OBJECTIVE: This systematic review aimed to investigate the neural substrates of apathy in PD by synthesizing evidence from structural and functional neuroimaging studies. METHODS: Following PRISMA guidelines, we searched PubMed, Embase, Web of Science, and Scopus for studies assessing apathy in PD through neuroimaging techniques, including structural MRI, resting-state functional MRI, diffusion tensor imaging, PET/SPECT molecular imaging, and connectivity analyses. Apathy severity was evaluated using validated clinical scales, and findings were compared between PD population with and without apathy, as well as healthy controls. RESULTS: Across 16 heterogeneous studies, recurrent findings implicated structural, functional, and molecular alterations within fronto-striatal-limbic networks. However, differences in apathy measures and neuroimaging modalities limit direct comparability across studies and do not support a single unified imaging signature of apathy in PD. Apathetic PD subjects consistently exhibited reduced gray matter volume and altered spontaneous activity in the dorsolateral, inferior frontal gyri, anterior cingulate cortex (ACC), and nucleus accumbens, alongside disrupted functional and structural connectivity in fronto-limbic tracts. These alterations were significantly correlated with apathy severity and, in several studies, remained detectable after accounting for depressive symptoms or global cognitive status. Furthermore, the available findings suggest that apathy in PD may involve heterogeneous neurobiological mechanisms, including variable involvement of frontal, striatal, and limbic circuits; however, current evidence remains insufficient to support firm transmitter-based subtyping or treatment stratification. CONCLUSIONS: Current evidence suggests that apathy in PD involves partially dissociable neural correlates across prefrontal, striatal, and limbic systems. However, methodological heterogeneity across studies limits strong generalization, and future harmonized longitudinal studies are needed to clarify which findings are robust and clinically meaningful.
16 June 2026
Read appraisal →Brain research
BDNF relates to prefrontal cortex activity in the context of physical exercise
Exercise has been shown to support brain health, cognitive function, and increase levels of brain-derived neurotrophic factor (BDNF). While BDNF is known to support the central nervous system through improved brain metabolism, vasculature, neurotransmission and synaptic plasticity, the association between exercise-induced changes in BDNF concentrations and exercise-related cognitive improvements is still unclear. This study investigated the relationship between exercise-induced changes in plasma BDNF (pBDNF) and serum BDNF (sBDNF), and haemodynamic indicators of prefrontal cortex function in sedentary adults. Participants (n = 23, female = 7) were randomized into intervention (12-week cycling programme) and control groups (no intervention). Participants completed V̇O2max tests to assess changes in fitness. pBDNF and sBDNF were measured pre- and post-V̇O2max testing, and a battery of executive function and memory tasks were also conducted, alongside functional near-infrared spectroscopy (fNIRS) to assess haemodynamic changes in the prefrontal cortex activity. Changes were assessed using the correlation-based signal improvement (CBSI) method. Results indicated that participants in the exercise intervention group did not show increased levels of resting-state s/pBDNF levels at the end of the intervention; however, they did exhibit a significant exercise-induced increase in sBDNF at week 12. This increase was correlated with changes in V̇O2max. Higher pBDNF levels and exercise-induced sBDNF were associated with a decrease in CBSI values in the frontopolar, dorsolateral and orbitofrontal prefrontal cortex during attention and inhibition tasks, but not during the memory task. These results suggest that increasing physical fitness can enhance BDNF transcription in response to acute bouts of exercise. This might, in turn, play a part in the modulation of neural function during executive tasks after acute exercise.
15 June 2026
Read appraisal →Journal of mathematical biology
Spiking neural models for decision-making tasks with learning.
In cognition, response times and choices in decision-making tasks are commonly modeled using Drift Diffusion Models (DDMs), which describe the accumulation of evidence for a decision as a stochastic process, specifically a Brownian motion, with the drift rate reflecting the strength of the evidence. In the same vein, the Poisson counter model describes the accumulation of evidence as discrete events whose counts over time are modeled as Poisson processes. This model has a spiking neurons interpretation as these processes are used to model neuronal activities. However, these models lack a learning mechanism and are limited to tasks where participants have prior knowledge of the categories. To bridge the gap between cognitive and biological models, we propose a biologically plausible Spiking Neural Network (SNN) model for decision-making that incorporates a learning mechanism and whose neurons activities are modeled by a multivariate Hawkes process. First, we show a coupling result between the DDM and the Poisson counter model, establishing that these two models provide similar categorizations and reaction times and that the DDM can be approximated by spiking Poisson neurons. To go further, we show that a particular DDM with correlated noise can be derived from a Hawkes network of spiking neurons governed by a local learning rule. In addition, we designed an online categorization task to evaluate the model predictions. This work provides a significant step toward integrating biologically relevant neural mechanisms into cognitive models, fostering a deeper understanding of the relationship between neural activity and behavior.
14 June 2026
Read appraisal →PloS one
Oral carbohydrate sensing enhances prefrontal cortex oxygenation, reduces perceived exertion, and improves high-intensity cycling performance: A randomized crossover trial
PURPOSE: Carbohydrate mouth rinsing (CHO-MR) and music listening (MUS) are non-metabolic interventions proposed to attenuate cognitive and perceptual fatigue during exercise. However, their comparative effects on prefrontal cortical oxygenation, executive function, and perceived exertion during high-intensity endurance performance remain unclear. This study examined the effects of CHO-MR and MUS on dorsolateral prefrontal cortex (DLPFC) oxygenation, cognitive performance, perceived exertion, and cycling performance during a 4-km time trial (TT). METHODS: Eleven trained cyclists (7 men, 4 women) completed a randomized, single-blind, crossover trial under three conditions: CHO-MR, MUS (120 beats/min), and placebo mouth rinse (PLA). Bilateral DLPFC oxygenation was assessed using functional near-infrared spectroscopy at rest, during Stroop testing, post-intervention, and post-TT. Stroop performance, rating of perceived exertion (RPE); every 500 m, completion time, power output, heart rate, and blood lactate were measured. Outcomes were analyzed using generalized estimating equations adjusted for period, sequence, and first-order carryover effects with robust standard errors. RESULTS: CHO-MR increased bilateral DLPFC oxygenation compared with MUS and PLA at post-intervention and post-TT time points (p < 0.05), with no differences at rest or baseline. Stroop performance was superior in CHO-MR relative to MUS and PLA following the TT (p < 0.001), without between-condition differences at earlier assessments. RPE was lower in CHO-MR than PLA across all intervals (p ≤ 0.01) and lower than MUS at mid-trial distances (p < 0.05). CHO-MR resulted in faster completion times, greater mean power output, and higher mean speed compared with MUS and/or PLA (p < 0.05), whereas peak power, heart rate, and blood lactate did not differ among conditions. Pooled change-score analyses demonstrated positive associations between bilateral ΔDLPFC oxygenation (p < 0.001) and between ΔDLPFC oxygenation and ΔStroop performance (p < 0.05). CONCLUSION: CHO-MR enhances bilateral DLPFC oxygenation, preserves executive function, and reduces perceived exertion during high-intensity endurance exercise, translating into improved performance without detectable peripheral metabolic alterations. These findings support a central neurocognitive mechanism underlying the ergogenic effects of oral carbohydrate sensing. TRIAL REGISTRATION: ClinicalTrials.gov NCT07099807.
14 May 2026
Read appraisal →Neuropsychologia
Sparsity and memory constraints interact with training sequence to bias learning of associative maps.
Cognitive maps support inference and planning by representing associations between experiences encoded in memory. These map-like representations are thought to carry information not only about directly observed links but also about longer paths. The ability to make judgments based on multi-step associations varies with one's experience in an environment and with changes in memory abilities across the lifespan. However, it remains unclear exactly how representations of associative structure are influenced by learning curricula and memory constraints. Prior studies have suggested a tradeoff: memory representations can either be more integrated to improve inference, or more separated to recall distinct direct associations. Whether overlapping associations are experienced nearby in time (interleaved) or spaced apart (blocked) can bias memory representations toward integration or separation. However, key recent findings about how blocked versus interleaved experience bias integration or separation have been inconsistent. Here, we introduce a computational framework that helps reconcile these apparent discrepancies. Using neural network simulations of three separate memory-guided inference tasks, we show that variations in memory capacity and the sparsity of neural codes interact with learning sequence to shape network representations. Specifically, blocked learning promotes integration when memory capacity is low, while interleaved learning promotes integration when memory capacity is high. Integration is especially likely to result from representations formed when neural codes are both sparse and distributed. These results offer a principled computational account of how flexible, map-like representations can arise from experience and suggest avenues for individualized memory interventions to improve inference, generalization, and planning.
10 May 2026
Read appraisal →International journal of psychophysiology : official journal of the International Organization of Psychophysiology
Task-evoked brain network dynamics underlying cognitive control using source-localized EEG and machine learning
OBJECTIVE: Functional connectivity (FC) coordinates brain activity during cognitive tasks, yet the influence of demographic variables and health factors on task-related dynamic FC remains underexplored. We examined FC within and between seven core brain networks: Default Mode (DMN), Ventral Attention (VAN), Somatomotor (SMN), Dorsal Attention (DAN), Frontoparietal (FPN), Visual (VIN), and Limbic (LIN) during a modified Eriksen flanker task. We aimed to identify distinct network configurations linked to varying cognitive control demands and assess the influence of demographic and health factors. METHODS: Secondary analyses of source-localized EEG data from 255 adults (mean age 38.97 ± 11.65 years; BMI 30.21 ± 9.29 kg/m2) were conducted. A hidden Markov model extracted spatiotemporal dynamics, quantifying within- and between-network correlations. Extreme Gradient Boosting with recursive feature elimination identified connectivity metrics predicting reaction time (RT) of high performers (>80% accuracy). Regression models assessed effects of age, sex, BMI, and income on task-related FC. RESULTS: Classification accurately predicted RT in both congruent (AUC = 0.81) and incongruent (AUC = 0.86) trials. Key FC features included VAN-LIN, VIN-VAN, and SMN-DAN for congruent trials; SMN-DMN, LIN-FPN, and VIN-DAN for incongruent trials. Age was linked to reduced FC, while higher BMI showed modest positive FC associations in LIN-FPN and VIN-DAN. Sex and income were not significant predictors of RT or FC. CONCLUSIONS: These findings reveal the relevance of specific dynamic network interactions in cognitive control and highlight the need to consider age and BMI as contributors to brain connectivity during task performance.
4 May 2026
Read appraisal →Psychological medicine
Characterizing executive dysfunctions in patients with schizo-obsessive comorbidity: comparing schizophrenia with obsessive-compulsive disorder
BACKGROUND: The term 'schizo-obsessive comorbidity (SOC)' is used to describe the presence of obsessive-compulsive symptoms or obsessive-compulsive disorder (OCD) in patients with schizophrenia (SOC). Recent studies have found overlapped executive dysfunctions in SCZ and OCD implicating shared pathophysiology. However, specific deficits in the components of executive function (EF) in patients with SOC remains unclear. METHODS: We recruited 37 patients with SOC, 68 patients with SCZ, 70 patients with OCD, and 59 healthy controls (HCs). All participants completed a battery of measures for EF components, namely initiation, sustained attention, online updating, switching, disinhibition, and planning. Apart from traditional group-mean analysis, we applied machine learning approaches to identify the unique patterns of EF among different clinical groups. RESULTS: The results showed that the three clinical groups could be distinguished from HCs. The feature importance analysis showed that, to classify clinical groups from HC, online updating was the core feature of SCZ patients, whereas disinhibition and online updating jointly determine classification between OCD patients and HC. In differentiating SOC from HC, online updating, planning, and disinhibition collectively served as key features. Machine learning algorithms classified SOC and OCD with acceptable performance but classified SOC and SCZ with lower performance. CONCLUSIONS: Deficits of EF are shared features among patients with SOC, SCZ, and OCD. However, the specific components of executive dysfunction in these clinical groups appeared distinct.
2 May 2026
Read appraisal →Behavioural brain research
Targeting the prefrontal-supplementary motor network with online and offline tDCS to modulate disgust: A single-blind and sham-controlled study
In certain individuals, disgust experience is abnormally high. This heightened sensitivity serves as a cognitive component that can lead to dysfunction in neural circuits related to emotion processing and psychological disorders such as contamination-based obsessive-compulsive disorder. In this single-blind study, 36 individuals with high disgust sensitivity were randomly assigned to three groups: online, offline, and sham. The online group received intensified and repetitive transcranial direct current stimulation (tDCS) while performing a disgust image watching task, whereas the offline and sham groups performed the task at two time points: immediately before (pre-test) and immediately after (post-test) the stimulation session. The offline group received active tDCS, whereas the sham group received placebo tDCS to control for the interval duration. Anodal and cathodal stimulation was applied to the left dorsolateral prefrontal cortex (l-DLPFC; F3) and pre-supplementary motor area (pre-SMA; FC2), respectively. The intervention consisted of 10 sessions of tDCS (2 sessions per day) over 5 days, with a stimulation intensity of 2 mA for 20 min and a 20-minute interval between sessions. The results showed a significant decrease in disgust experience in the online group compared to the offline group, while there was no significant difference between the online and sham groups. Although these findings suggest potential effects of tDCS combined with a task, further studies are necessary to confirm its efficacy.
1 May 2026
Read appraisal →Neurology
Association of MRI-Visible Perivascular Spaces With Longitudinal Cognitive Decline Over a Decade
BACKGROUND AND OBJECTIVES: Cerebral small vessel disease (SVD) is the most common vascular contributor to dementia. SVD markers often coexist, contributing to difficulty assessing their independent contributions to cognitive domains. MRI-visible perivascular spaces (PVSs) are an emerging SVD marker visualized by MRI. We previously showed that basal ganglia PVSs cross-sectionally contribute to worse cognition, independent of other SVD markers. To further characterize the clinical relevance of PVS, we studied its role as a unique SVD marker of longitudinal cognitive decline. METHODS: Participants without stroke or dementia were included in the Vanderbilt Memory and Aging Project, a longitudinal observational cohort study based in Nashville, TN. Participants completed 3T MRI at study entry to measure SVD burden (PVS volume fraction, white matter hyperintensities volume, lacune counts, and cerebral microbleeds counts). PVS volumes were segmented using a deep learning algorithm. Participants underwent comprehensive serial neuropsychological testing over an 11-year follow-up period (mean follow-up = 4.9 ± 3.1 years). Each SVD marker was related to longitudinal neuropsychological performances using a linear mixed-effects model adjusting for age, sex, race/ethnicity, education, baseline cognitive status, apolipoprotein E-ε4 presence, Framingham Stroke Risk Profile, and intracranial volume. Head-to-head comparisons simultaneously tested multiple statistically significant SVD markers. RESULTS: Among participants (n = 750, age 68 ± 9 years, 52% female), higher basal ganglia PVS burden at baseline was associated with worse longitudinal performances in Boston Naming Test (β = -29.63; 95% CI -56.66 to -2.60), Animal Naming (β = -33.09; 95% CI -65.66 to -0.51), Wechsler Adult Intelligence Scale IV Coding (β = -86.36; 95% CI -150.9 to -21.82), executive function composite (β = -9.51; 95% CI -14.33 to -4.68), Hooper Visual Organization Test (β = -26.06; 95% CI -50.53 to -1.59), and episodic memory composite (β = -7.05; 95% CI -11.9 to -2.21). In head-to-head comparisons, basal ganglia PVS remained independent associations with executive function composite (β = -7.47; 95% CI -12.84 to -2.10) and Hooper Visual Organization Test (β = -22.11; 95% CI -42.38 to -1.85). DISCUSSION: Basal ganglia PVS burden independently contributes to worse longitudinal executive function and visuospatial skills independent of other SVD markers, highlighting PVS as an emerging marker of domain-specific cognitive decline in aging. Although causation cannot be established, findings further support PVS as a vascular contributor to deep brain structure damage underlying cognitive decline over time.
28 Apr 2026
Read appraisal →Brain topography
Effects of Sleep Deprivation and Neuromodulatory Interventions on Olfactory-Related Brain Function
Sleep deprivation (SD) has been shown to impair sensory functions, including olfactory processing, which may be related to changes in brain activity. This study aims to investigate the effects of SD, short-term sleep recovery (RS), and neuromodulation interventions on the functional activity of olfactory-related brain regions. We hypothesize that these interventions could restore some of the neural activity disrupted by SD, particularly in areas associated with olfactory processing. In this study, 90 male participants first underwent a baseline sleep period of at least 8 h, followed by 36 h of sleep deprivation (SD36h). The participants were randomly assigned to four groups: control group (CON, n = 30, age 20.8 ± 1.6 years), magnetic stimulation group (MS, n = 30, age 21.3 ± 1.6 years), electrical stimulation group (ES, n = 17, age 20.3 ± 0.7 years), and combined stimulation group (CS, n = 13, age 20.8 ± 1.0 years). After SD36h, participants underwent an 8-hour restorative sleep period, before and after completing the 8-hour RS period, additionally received 30 min of MS(MS was applied over the occipital region using a rhythmic gamma magnetic-field device), ES(ES was delivered as tDCS targeting the left DLPFC, anode at F3, cathode at FP1, 2 mA), and CS. All participants underwent three resting-state functional MRI (fMRI) scans: at baseline, after sleep deprivation, and after intervention. This experiment analyzed the Amplitude of Low-Frequency Fluctuations (ALFF) and functional connectivity (FC) in olfactory-related regions. ALFF and FC were analyzed using a group (CON/MS/ES/CS) × time (baseline/SD36h/RS) repeated-measures ANOVA with FDR correction, followed by FDR-corrected post-hoc tests for significant main effects. No group × time interaction survived FDR correction for ALFF or FC. For ALFF, a main group effect was observed in Hippocampus_L and Olfactory_R, with the ES (and CS in Olfactory_R) showing lower ALFF than CON and/or MS. For time effects, ALFF in Hippocampus_L was reduced at SD36h and remained lower after RS relative to baseline, while Olfactory_L ALFF was lower after RS than baseline. For FC, main group effects were found in Hippocampus_L and Fusiform_R, with ES and CS showing higher FC than CON and MS. A main time effect cluster with a peak in Cingulum_Mid_R showed reduced FC after SD36h and increased FC after RS, returning to a level not significantly different from baseline. SD disrupts the ALFF and FC in olfactory-related brain regions, while RS can partially restore these functions. Short-term neuromodulation interventions, such as electrical stimulation and combined stimulation, may improve network-level connectivity but are insufficient to fully restore local neural activity.
26 Apr 2026
Read appraisal →PLoS computational biology
'Backpropagation and the brain' realized in cortical error neuron microcircuits
Neural responses to mismatches between expected and actual stimuli have been widely reported across different species. How does the brain use such error signals for learning? While global error signals can be useful, their ability to learn complex computation at the scale observed in the brain is lacking. In comparison, more local, neuron-specific error signals enable superior performance, but their computation and propagation remain unclear. Motivated by the breakthrough of deep learning, this has inspired the 'backpropagation and the brain' hypothesis, i.e., that the brain implements a form of the error backpropagation algorithm. In this work, we introduce a biologically motivated, multi-area cortical microcircuit model, implementing error backpropagation under consideration of recent physiological evidence. We model populations of cortical pyramidal cells acting as representation and error neurons, with bio-plausible local and inter-area connectivity, guided by experimental observations of connectivity of the primate visual cortex. In our model, all information transfer is biologically motivated, inference and learning occur without phases, and network dynamics demonstrably approximate those of error backpropagation. We show the capabilities of our model on a wide range of benchmarks, and compare to other models, such as dendritic hierarchical predictive coding. In particular, our model addresses shortcomings of other theories in terms of scalability to many cortical areas. Finally, we make concrete predictions, which differentiate it from other theories, and which can be tested experimentally.
19 Apr 2026
Read appraisal →Neuroscience
Effects of single/dual-target transcranial alternating current stimulation on working memory in healthy individuals
This study compared the effects of single- and dual-target high-definition transcranial alternating current stimulation(tACS) on working memory and explored the associated neural mechanisms using functional near-infrared spectroscopy(fNIRS). In a randomized, single-blind, sham-controlled design, 45 healthy participants received either dual-target stimulation, single-target stimulation, or sham stimulation, before and after which working memory performance was performed using n-back tasks with simultaneous fNIRS recording. Behavioral results showed that dual-target stimulation improved working memory performance, particularly under higher cognitive load, and yielded a greater proportion of responders when compared with sham stimulation, whereas single-target stimulation induced relatively limited effects. fNIRS findings indicated that dual-target stimulation modulated task-related cortical activation in prefrontal and temporal regions and enhanced both intra- and inter-hemispheric functional connectivity. Moreover, improvements in task accuracy were associated with strengthened functional connectivity within the frontoparietal and default mode networks. These findings suggest that dual-target tACS may be a promising strategy for enhancing working memory and highlight the importance of network-level modulation in cognitive enhancement.
6 Apr 2026
Read appraisal →