SEARCH · Search PubMed

Results for “Student learning”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

182 records · Page 4Linked to original sources

PubMed2026

Revealing potential biomarkers and metabolic mechanisms of ovarian aging in hens during late laying period based on machine learning and metabolomics.

Ovarian function decline during the late laying period represents a major bottleneck for the economic efficiency of the global poultry industry. However, the underlying metabolic mechanisms and reliable early-warning biomarkers for ovarian aging remain poorly understood. In this study, we performed the first untargeted LC-MS/MS metabolomics analysis of ovarian tissues from Taihe silky fowls at peak laying (30 weeks) and late laying (50 weeks) stages, and employed an ensemble machine learning strategy integrating LASSO, random forest, and support vector machine (SVM) algorithms to identify high-confidence core biomarkers of ovarian aging. Gene expression analysis was further conducted to validate the potential molecular mechanisms. Our results showed that the metabolic profiles of ovarian tissues differed significantly between the two groups. A total of 6 core biomarkers were identified, 4 of which were long-chain acylcarnitines. Mechanistic analysis revealed that downregulation of key genes in the carnitine shuttle system led to impaired mitochondrial fatty acid β-oxidation, which in turn triggered excessive oxidative stress and compromised ovarian endocrine function. In conclusion, this study identifies long-chain acylcarnitines as potential metabolic biomarkers for ovarian aging in Taihe silky fowls. These findings provide novel insights into the metabolic basis of poultry ovarian aging and lay a theoretical foundation for the precise regulation of reproductive performance in indigenous poultry breeds.

Animals↗

PubMed2026

Improving insurance deduction identification: a hybrid artificial intelligence model using machine learning and expert systems.

PURPOSE: Financial challenges in healthcare systems worldwide, especially in low- and middle-income countries like Iran, have increased hospitals' reliance on insurance reimbursements. Unrecognized insurance deductions often cause severe financial shortages, making efficient deduction management crucial. This study aimed to design a hybrid intelligent system for identifying and predicting insurance deductions by combining machine learning and expert system frameworks. DESIGN/METHODOLOGY/APPROACH: A mixed-methods design was applied in four stages. First, a scoping review identified the causes and patterns of insurance deductions. Second, interviews with 15 insurance experts produced a validated checklist and a dataset from inpatient billing records. Third, using the CRISP-DM methodology, machine learning algorithms were developed and tested in SPSS Modeler alongside a fuzzy expert system developed in MATLAB. Finally, the model was validated using the holdout method. FINDINGS: Four categories of deduction drivers were identified: service provision, registration errors, document submission issues, and revenue conversion processes. The CHAID decision tree outperformed other algorithms with a 99% precision rate and the lowest Mean Absolute Error (9.43). A brief assessment of potential overfitting was conducted to ensure that the CHAID model's high accuracy was interpreted cautiously and supported by the validation results. The fuzzy expert system with validated rules was adaptable for deduction classification, especially for cases unsuitable for quantitative modeling. ORIGINALITY/VALUE: The hybrid model improves detection and prevention of deductions, offering actionable insights for hospital administrators, insurers, and policymakers. Its implementation can enhance hospital information systems, streamline claims processing, and optimize revenue management amid financial constraints.

Machine Learning↗

PubMed2026

Response-optimised training improves learning of a complex motor task and closely related motor tasks.

Regular physical exercise is essential for promoting healthy aging and longevity. In older adults with varying physical and cognitive decline, optimising exercise interventions is crucial to maximise benefits. A promising approach to achieve this goal is by adjusting task demands to individual abilities in turn preventing over- or underloading their abilities. In the field of motor learning, it is currently unclear whether such an optimised training improves not only performance on the trained task but also transfers to untrained motor and cognitive tasks. We conducted a randomized, single-blinded, 6-week dynamic balance training (DBT) with healthy older adults (n = 30). Training was tailored to individual balance ability. Participants were assigned to either suboptimal (high or low difficulty) or optimal (moderate difficulty) training groups. Transfer effects were assessed via cognitive tasks (memory and executive) and motor tasks (untrained DBT variations and other balance tasks) measured pre-, mid- and post-intervention. Multivariate longitudinal statistical analysis showed higher performance gains in the optimal training group in three out of six sessions compared to the suboptimal groups, especially under testing conditions with high task demands. The optimal group also showed greater improvements in near motor transfer tasks mid- and post-intervention, while no significant differences were observed in the cognitive tasks. Within-group DBT learning positively correlated with transfer gains, highlighting the role of training response in achieving transfer. In conclusion, optimised task difficulty in balance training enhances both task-specific performance and related motor skills, supporting the use of personalised interventions to maintain function and independence in older adults.

Humans↗

PubMed2026

Predicting ACL injury risk in athletes: A systematic review of machine learning-based models.

BACKGROUND: Early ACL injury risk identification in athletes is essential. This systematic review examines machine learning (ML) models for predicting ACL injuries, evaluating their methodological quality, performance, and reliability. METHOD: A comprehensive electronic search was conducted across PubMed, Scopus, Web of Science, and IEEE Xplore databases, supplemented by Google Scholar for grey literature, covering articles published between January 1, 2015, and August 30, 2025. Eligible studies were appraised using the Prediction Model Study Risk of Bias Assessment Tool (PROBAST) for methodological quality and risk of bias, and the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) guidelines for quality of evidence. RESULTS: Ten studies were included. PROBAST showed eight studies had moderate risk of bias and two low risk. TRIPOD found only two studies met quality criteria. ML models included logistic regression (n = 5), support vector machines (n = 4), k-nearest neighbor (n = 3), decision trees (n = 3), random forests (n = 5), neural networks (n = 2), linear discriminant analysis (n = 1), and pre-trained CNNs (n = 1). AUC ranged from 0.63 to 0.98. Accuracy (reported in six studies) ranged from 26% to 95%; however, these values should be interpreted with caution due to the absence of confidence intervals, lack of class imbalance handling, and limited external validation across studies. Tree-based ensemble methods such as random forest achieved competitive accuracy (74-86%), while SVM, a non-ensemble classifier, reported accuracy ranging from 71% to 95%; however, the highest values were obtained in studies with notably small sample sizes (n = 12 to n = 39), raising concerns about overfitting and generalizability. CONCLUSION: Current ML algorithms show promise for identifying athletes at high ACL injury risk and detecting relevant risk factors. Although study quality was generally satisfactory, future research should prioritize external validation and model interpretability to support clinical translation.

Humans↗

PubMed2026

Not just when, but how: An exploratory dual-control approach to video feedback in motor learning.

The present study provides exploratory evidence for a novel dual-control paradigm. It examines whether combining temporal over video feedback timing with learner-controlled interactive playback functions (pause, slow-motion, rewind) would enhance motor skill acquisition beyond temporal autonomy alone. Sixty-four novice adults were randomly assigned to one of four conditions: Full Control (self-controlled timing + interactive replay), Partial Control (self-controlled timing + non-interactive replay), Yoked Full Control (externally controlled timing + interactive replay), or Yoked Partial Control (externally controlled timing + non-interactive replay). Motor accuracy (Radial Error), movement consistency (Bivariate Variable Error), technical execution, and self-efficacy were assessed at pre-test, 24-h retention, and 72-h retention following two acquisition sessions on a dart-throwing task (120 trials total). The Full Control group demonstrated the greatest and most durable learning gains across all outcomes. The Group × Time interaction was significant across all dependent variables (η2ₚ ranging from 0.140 to 0.234), with Full Control demonstrating superior retention at both 24 and 72 h relative to other groups (though differences relative to Partial Control were more pronounced at 72-h retention). Critically, the Yoked Full Control group showed comparatively weaker outcomes despite access to the same interactive playback functions. These findings suggest that interactive video tools may be most useful when learners can regulate both when feedback is accessed and how it is inspected. Theoretical and practical implications for the design of learner-centered video feedback systems are discussed.

Humans↗

PubMed2026

Machine learning-ready genomic biomarkers: ATF3 polymorphisms predict postoperative analgesic demand through AI-compatible phenotyping.

PURPOSE: To determine whether ATF3 polymorphisms can serve as genetic biomarkers for machine learning-based precision analgesia by establishing a genotype-phenotype association suitable for predictive modeling of postoperative opioid requirements. METHODS: In a prospective cohort of 167 adults undergoing abdominal surgery, ATF3 SNPs rs3122721 and rs3125293 were genotyped. A structured dataset architecture was developed to represent genetic profiles as input features for supervised learning models, enabling translational analysis of genotype‑dependent opioid consumption over 72 h. RESULTS: Patients with homozygous genotypes of the ATF3 SNPs had significantly higher opioid requirements than non‑carriers, despite reporting similar subjective pain scores. This consistent genotype‑dependent pattern provided a clinically relevant phenotype suitable for integration into predictive algorithms. CONCLUSION: ATF3 genotyping offers a promising biomarker for computationally informed precision analgesia. By linking genomic variability to clinically meaningful outcomes within a structured clinical and genomic framework, this approach supports the future development of risk-stratified clinical decision-support systems to optimize postoperative pain management.Trial registration ChiCTR1900021991, registered 30 April 2019. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at https://doi.org/10.1007/s13755-026-00480-9.

ATF3↗

PubMed2026

Predicting training outcomes for developmental dyslexia from EEG data.

Developmental dyslexia (DD) is characterised by lower-than-average reading abilities and is diagnosed in approximately 10% of individuals. The societal barriers may limit professional fulfilment and psychological wellbeing of individuals with DD, calling for the development of effective interventions to counteract them. As DD is associated with challenges in both phonological and visuo-attentional domains, different longitudinal training approaches were developed to strengthen them. However, they require a considerable amount of personal, social and economic resources and the outcomes may vary depending on individual differences in behavioural and neurophysiological functionality. Hence, predicting training outcomes might help in developing personalised treatment protocols and optimising the use of resources. In the present work we applied machine learning to resting-state EEG to predict longitudinal training outcomes in adults with DD enrolled in a randomized clinical trial. In particular, one group received a visuo-attentional training combined with transcranial alternating current stimulation (tACS), another group received visuo-attentional training with sham/placebo stimulation, and the third group received a phonological training with sham/placebo stimulation. The improvement in text reading speed was associated with spectral power in low-beta and individual frequencies in the alpha (IAF) and beta (IBF) bands, while the improvement in pseudoword reading was associated with IBF. The findings highlight the potential of capturing neural markers of treatment responsiveness in DD. Future studies should focus on the generalisability of predictive models to real-world settings, while investigating whether specific EEG markers predict responsiveness to distinct remediation protocols, thus supporting the development of personalised interventions.

Humans↗

PubMed2026

The role of simulator immersion on learning and transfer of decision-making skill in sport.

Virtual reality has become popular in sport and other domains because it can immerse the user within a sporting context and solve logistical problems for additional off-field training. There is limited evidence, however, of whether immersion is crucial for learning and transfer. This study compared training of decision-making skill between 360-degree video virtual reality (360VR) and two-dimensional video. Twenty-eight Australian Rules Football players were randomly assigned to one of three training groups: 360VR, two-dimensional video, and control. Across four weeks, participants in the training groups were exposed to decision-making scenarios consisting of visual, contextual and auditory cues. Performance was assessed pre- and post-training with virtual reality and field-based decision-making tests. Results indicated that the two-dimensional video training group showed significantly superior decision-making in the field-based transfer test compared to 360VR and control groups post intervention. There was also indication that two-dimensional video training was superior to the control post intervention in the virtual reality test. Findings indicate that immersion created in virtual reality is not an underpinning mechanism for learning and transfer, rather the use of perceptual information is crucial. 360VR may facilitate uptake through engagement, but two-dimensional video is adequate for learning and transfer of decision-making to the field.

Humans↗

PubMed2026

Mul-PheG2P: decoupled learning and prediction-space fusion enables robust and interpretable multi-phenotype genomic prediction.

Genomic prediction of multiple phenotypes is crucial in modern plant breeding; however, existing methods struggle with negative transfer and lack interpretability, particularly across high-dimensional small-sample data and diverse species. To address this, we propose Mul-PheG2P, a novel paradigm based on decoupled learning and predictive space fusion. It employs a two-stage design: first training phenotype-specific encoders using genetic data, then decoupling phenotype-specific learning from cross-phenotype aggregation via an interpretable prediction layer. Mul-PheG2P outperforms existing methods across diverse crop datasets, including maize (Zea mays), wheat (Triticum aestivum), and tomato (Solanum lycopersicum). It provides a multi-scale interpretability chain: at the macro level, it quantifies phenotypic contributions via attention-based weighting; at the micro level, Integrated Gradients reveal the genetic basis of predictions. Notably, the model successfully identified the CCT (CONSTANS, CO-like, and TOC) motif regulating photoperiodism and the SQUAMOSA (SQUAMOSA promoter binding protein) promoter for inflorescence development, confirming its ability to capture functional biological mechanisms. These results highlight the high performance and interpretability of Mul-PheG2P, showcasing its value for low-cost, large-scale screening to advance precision breeding.

Phenotype↗

PubMed2026

Future promise, current clinical ambiguity: a systematic review of machine learning algorithm outputs predicting risk of cardiovascular disease.

OBJECTIVE: To examine whether the outputs of machine learning algorithms designed to predict risk of cardiovascular disease (CVD) address known deficiencies of the Framingham Risk Score (FRS) and improve risk estimates. METHODS: For this critical review, Medline, Embase and IEEE were searched from inception to 1 January 2025. Included were studies describing machine learning algorithms designed to specifically compare output of cardiovascular risk assessment with the FRS. Commentaries, letters, unpublished work or non-peer-reviewed papers were excluded.Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, two reviewers screened titles and abstracts independently, then populated a purpose-built data extraction form. A subsequent qualitative thematic analysis focused on algorithms' strengths, added value, potential harms, unintended consequences and equity implications.The main outcome assessed was whether, among healthy adults, the algorithm improved CVD risk prediction relative to the FRS. RESULTS: Of 707 studies retrieved, 29 met inclusion criteria. 23 reported improved predictive ability relative to the FRS. Most datasets and/or medical records used included sociodemographic predictors of CVD not included among FRS inputs. Some added costly diagnostic tests like CT angiography to FRS screening indicators. When they were defined, inputs and outcomes such as hypertension or myocardial infarction did not always adhere to FRS values. Statistical significance was generally taken as a proxy for clinical significance. Some algorithms overestimated the number at risk compared with the FRS without discussing whether that larger proportion might be at risk of overdiagnosis rather than CVD, while a few decreased the proportion found to be at risk. CONCLUSIONS: Use of artificial intelligence to improve accuracy of risk assessment for CVD demonstrates the technological capacity to merge known sociodemographic predictors with biologic variables and examine non-linear interactions among these. Still needed to achieve patient benefit is clinical insight, adherence to screening principles and cost-benefit assessment of inputs selected.

Humans↗

PubMed2026

Meta-PseU: A meta-classifier for robust prediction of RNA pseudouridine modification sites from long sequences.

BACKGROUND AND OBJECTIVES: Pseudouridine (Ψ) represents one of the most abundant and conserved RNA modifications. Ψ provides an additional hydrogen-bond donor that enhances RNA structural stability and modulates translation. It participates in diverse biological processes, including RNA-protein interactions, splicing, translational control, and stress responses. Aberrant pseudouridylation is implicated in cancer, neurodegenerative disorders, and autoimmune diseases. Despite its biological importance, experimental identification of Ψ sites remains time-consuming and costly, limiting the feasibility of transcriptome-wide profiling. Computational approaches have therefore become essential complements to experimental techniques. However, state-of-the-art machine-learning and deep-learning predictors often suffer from limited generalizability due to small training datasets. To overcome these issues, we aim at constructing new long-sequence datasets and developing a novel Ψ site predictor. METHODS: New long-sequence datasets were constructed as benchmarks for RNA Ψ-site prediction. The Ψ modification sites in RMBase 3.0 were mapped to the reference genomes across three species of human, mouse, and yeast, and the RNA sequences with a length of 201 were generated by extending the upstream and downstream from the mapped, central sites. To eliminate sequence redundancy, the sequences were clustered using CD-HIT with a 70% sequence identity threshold. We developed Meta-PseU, a logistic regression-based meta-classifier that considered 118 machine learning and deep learning classifiers. The datasets and programs are freely accessible at https://github.com/kuratahiroyuki/MetaPseU. RESULTS: By optimizing model configuration, we proposed the Meta-PseU model stacking 32 machine learning and deep learning classifiers out of 118 classifiers. Meta-PseU substantially improved model generalizability, overcoming a key limitation of existing approaches. It greatly outperformed state-of-the-art predictors and achieved increasing accuracy with increasing sequence length. CONCLUSIONS: Long-sequence datasets were newly constructed as benchmarks for RNA Ψ-site prediction. Meta-PseU offers a new framework for robust Ψ-site identification by using long sequences.

Pseudouridine↗

PubMed2026

Diagnostic performance of machine learning models versus established risk stratification for intracranial aneurysm rupture: a systematic review and bivariate meta-analysis.

BACKGROUND: Machine learning (ML) models have been proposed to improve the discrimination of intracranial aneurysm rupture status beyond established clinical risk stratification tools. However, reported performance is heterogeneous and the relative contribution of model architecture and feature dominance remains unclear. METHODS: We performed a Preferred Reporting Items for Systematic Reviews and Meta-Analyses-diagnostic test accuracy systematic review and diagnostic meta-analysis of studies evaluating ML models for intracranial aneurysm rupture discrimination. PubMed, Embase and CENTRAL were searched to February 2026. Sensitivity and specificity were pooled using a bivariate random-effects model, with summary receiver operating characteristic curves generated across training, internal testing and external validation datasets. Models were compared with regression-based approaches and Population, Hypertension, Age, Size of aneurysm, Earlier subarachnoid haemorrhage, Site of aneurysm (PHASES) scores. Subgroup and meta-regression analyses explored associations between algorithm family and feature domain. RESULTS: Sixty-two retrospective cohorts (29 709 patients 209 models) met the inclusion criteria. In training datasets, pooled sensitivity and specificity for ML were 0.81 (95% CI 0.75 to 0.85) and 0.83 (0.80-0.86), with an area under the curve (AUC) of 0.878, exceeding PHASES (AUC 0.667). In testing datasets, ML retained higher discrimination (AUC 0.837) than regression models (0.806) and PHASES (0.646). In external validation, sensitivity was preserved (0.82), but specificity declined (0.66). Deep learning demonstrated the highest AUCs (training and testing). Incorporation of haemodynamic or radiomic features improved pooled discrimination relative to morphology alone. Evidence of small-study effects and mostly unclear Prediction Model Risk Of Bias Assessment Tool ratings were observed. CONCLUSIONS: ML approaches demonstrate higher pooled discrimination for aneurysm rupture status than conventional risk scores in retrospective datasets, but reduced external validation specificity and heterogeneity limit confidence for clinical translation. Prospective, externally validated, calibrated models are required before integration into routine cerebrovascular risk stratification.

Humans↗

PubMed2026

Data-centric, robust, and explainable multimodal deep learning for clinical decision support: A systematic review.

PURPOSE: Multimodal deep learning is increasingly proposed for clinical decision support (CDS) under a "data-centric" framing that prioritizes label quality, missing-modality robustness, distribution shift, calibration, and explainability. Prior reviews have examined multimodal medical AI, CDS, and data-centric methods separately, but none address their intersection. We mapped the modalities, fusion strategies, and data-centric and explainability techniques used in this recent literature, quantified how often each is implemented rather than merely mentioned, assessed deployment-relevant evidence (external validation, clinical-outcome measurement, equity), and formally appraised study-level risk of bias. METHODS: Following the PRISMA 2020 statement (PROSPERO CRD420261427815; registered retrospectively), we screened 150 records and included primary, clinical, multimodal studies that applied machine or deep learning to a decision-support task and reported at least one quantitative result. Two reviewers screened and extracted data with consensus adjudication. Each study was coded against pre-specified operational definitions, separating implemented or empirically evaluated techniques from those only mentioned. Study-level risk of bias was assessed with PROBAST + AI. Synthesis was narrative. RESULTS: Thirty-one studies met inclusion; 30 (97%) were published between 2024 and 2026, with a median of three modalities (range 2-6), most commonly structured EHR (71%) and imaging (39%). Data-centric techniques were frequently reported (74-84% across label-noise, distribution-shift, calibration, missing-modality and class-imbalance handling; equity 61%). However, external validation was reported in only 4/31 studies (13%), a clinical or provider outcome in 3/31 (10%), and no study reported routine deployment. Overall risk of bias was high in 27/31 studies (87%), driven by the analysis domain. CONCLUSION: Within this recent, self-selected slice of the field, technical robustness and explainability techniques are widely reported but rarely validated out-of-distribution or against clinical outcomes, and the underlying evidence is at high risk of bias. Progress requires external multi-site validation, clinical-outcome measurement, formal bias appraisal, and adherence to AI reporting standards (e.g., TRIPOD + AI) before deployment can be justified.

Deep Learning↗

PubMed2026

Clinical Efficacy and Learning Curve of Far-Lateral Approach (FLA) in Uni-Portal Non-Coaxial Spinal Endoscopic Surgery (UNSES) in the Treatment of Lumbar Degenerative Diseases: A Prospective Study.

BACKGROUND: Uniportal non-coaxial spinal endoscopic surgery (UNSES) via far-lateral approach (FLA) is an innovative minimally invasive procedure for lumbar degenerative diseases, particularly far-lateral disc herniation and foraminal stenosis. However, complex lateral lumbar anatomy and strict endoscope-instrument coordination create a distinct learning curve that may compromise early surgical efficiency and safety. This study aimed to evaluate the efficacy and safety, quantify the learning curve, and to provide clinical guidance for the standardized promotion and application of this technology. METHODS: A total of 40 consecutive patients with lumbar degenerative diseases who underwent UNSES via FLA by a single surgeon between January 2025 and December 2025 were included. All data were analyzed using SPSS 26.0 statistical software (IBM, USA). Primary outcomes included operation time, blood loss, fluoroscopy frequency, and intraoperative complication rate. Secondary outcomes were VAS, ODI, and modified Macnab criteria at 1, 3, and 6 months postoperatively. The learning curve and the inflection point of the learning curve was determined using cumulative sum (CUSUM) analysis. The differences in clinical indicators between early and proficient stage were compared. RESULT: Operation time, blood loss, and fluoroscopy times decreased significantly with case accumulation (p < 0.05). CUSUM identified an inflection point at the 16th case, after which operation time stabilized at (55.3 ± 8.6) min, much shorter than the early phase (89.5 ± 10.3) min (p < 0.001). Before the 16th case, the curve was in an upward trend; after the 16th case, the curve tended to be flat, indicating the proficiency stage. Postoperative VAS and ODI improved significantly than those before surgery at each follow-up time (p < 0.05). There was no significant difference in postoperative VAS score and ODI between the two groups at each follow-up time point (p > 0.05). The total complication rate was 12.5% (5/40), were cured by conservative treatment. The total excellent-good rate was 90.0% (36/40). L5/S1 and Bertolotti's syndrome were independent factors affecting the learning curve. CONCLUSION: UNSES via FLA is a safe and effective minimally invasive technique for treating complex lumbar degenerative diseases. It has a certain learning curve, and the inflection point is about the 16th case. After mastering the key techniques such as anatomical positioning, endoscopic manipulation and hemostasis, the surgeon can gradually reach the proficiency stage, with significantly improved surgical efficiency and clinical efficacy, and controllable complications. This study provides a theoretical basis for the clinical training and technology promotion of UNSES via FLA.

Humans↗

PubMed2026

Reliability-aware hierarchical learning for Chagas disease screening from 12-lead ECGs: tackling label uncertainty and class imbalance.

Objective.Chagas disease, a neglected tropical disease (NTD) with significant cardiovascular impact, remains underdiagnosed in resource-limited regions. Electrocardiogram (ECG) screening offers a low-cost tool for detecting cardiac involvement, yet algorithm development is challenged by label noise, data scarcity, and the latent nature of infection. This study proposes a robust ECG-based screening framework that explicitly addresses these constraints.Approach.We introduce aReliability-Aware Hierarchical Learningstrategy that calibrates supervision according to data provenance, prioritizing serology-confirmed labels over noisy self-reports. To mitigate data scarcity, we compare a specialized convolutional neural network (CNN) trained from scratch with a transfer learning approach based on a Spatio-Temporal ECG foundation Model (FM). Performance is evaluated across varying data scales, and the representation structure is analyzed to interpret model behavior.Main results.On the official hidden test set of the George B. Moody PhysioNet/Computing in Cardiology Challenge 2025, our approach achieved a Challenge Score of 0.163. We observe that while the specialized CNN performs competitively in data-rich regimes, the FM exhibits superior robustness in extreme low-resource settings. Furthermore, performance reaches a plateau imposed by underlying disease physiology. Bimodal score distributions suggest that models distinguish established cardiomyopathy from indeterminate infection, which remains electrophysiologically indistinguishable from healthy controls.Significance.These findings clarify both the potential and intrinsic limits of ECG-based AI screening for NTD-associated cardiac involvement. Reliability-aware supervision and data-efficient transfer learning provide a practical framework toward scalable and clinically meaningful ECG screening systems in resource-constrained environments.

Humans↗

PubMed2026

Dynamic evolution of chaperone-mediated autophagy is associated with tumor microenvironment remodeling and prognostic stratification in lung adenocarcinoma: insights from single-cell transcriptomics, ensemble machine learning, and experimental validation.

BACKGROUND: Lung adenocarcinoma (LUAD) shows prognostic heterogeneity, and tumor-node-metastasis (TNM) staging is limited for individualized management. Chaperone-mediated autophagy (CMA) maintains proteostasis, but its role during adenocarcinoma in situ (AIS)-minimally invasive adenocarcinoma (MIA)-invasive adenocarcinoma (IAC) progression remains unclear. METHODS: Single-cell RNA sequencing (scRNA-seq) data from GSE189357 and bulk transcriptomes from The Cancer Genome Atlas (TCGA)-LUAD and Gene Expression Omnibus (GEO) cohorts were integrated. CMA activity, cell-cell communication, weighted gene co-expression network analysis (WGCNA), tumor-normal differential expression, machine-learning survival modeling, tumor microenvironment (TME) features, drug sensitivity, and EPC1 function were analyzed. RESULTS: CMA-high tumor epithelial cells increased from AIS (58.1%) to MIA (65.7%) but declined in IAC (44.4%; p < 0.001). CMA-low cells preferentially received fibroblast-derived extracellular matrix cues. A CMA-negatively correlated module identified 69 core genes. Random survival forest (RSF) performed best among 117 machine-learning combinations (mean concordance index > 0.873). High-risk patients had worse survival across cohorts, and the risk score was independently associated with overall survival (hazard ratio = 16.013, 95% confidence interval: 9.579-26.768, p < 0.001). High-risk tumors showed proliferative activation and M0 macrophage enrichment, whereas low-risk tumors showed stronger immune-related signaling. EPC1 overexpression suppressed malignant phenotypes in A549 cells. CONCLUSION: CMA dynamics are associated with stromal and immune remodeling during LUAD progression. A CMA-based model provides robust prognostic stratification and may offer a basis for future TME-guided studies.

Chaperone-mediated autophagy↗

PubMed2026

Diagnostic Performance of Machine Learning for Systemic Lupus Erythematosus: Systematic Review and Meta-Analysis.

BACKGROUND: Early and accurate diagnosis of systemic lupus erythematosus (SLE) and its organ involvement is essential. Previous reviews of machine learning (ML) in SLE combined heterogeneous tasks and validation strategies and may have overinterpreted model performance. OBJECTIVE: This study evaluated the diagnostic performance of ML and deep learning (DL) models for 3 clinically distinct SLE-related tasks: SLE classification or diagnosis, lupus nephritis (LN) diagnosis, and neuropsychiatric systemic lupus erythematosus (NPSLE) discrimination. We also assessed methodological quality and certainty of evidence. METHODS: PubMed, Embase, Cochrane Library, Web of Science, and IEEE Xplore were searched from January 2014 to April 2026. Eligible peer-reviewed diagnostic accuracy studies developed or validated ML or DL models for 1 of the 3 prespecified tasks, used an accepted reference standard, and provided data for a 2×2 contingency table. Bivariate random-effects meta-analyses with the Hartung-Knapp-Sidik-Jonkman adjustment were used to pool sensitivity and specificity. We reported 95% prediction intervals (PIs), assessed risk of bias using the Quality Assessment of Diagnostic Accuracy Studies for Artificial Intelligence tool (QUADAS-AI; Viknesh Sounderajah [Imperial College London]), and evaluated certainty of evidence using the Grading of Recommendations Assessment, Development, and Evaluation framework for diagnostic test accuracy. RESULTS: Twenty-nine studies were included: 17 for SLE classification, 5 for LN diagnosis, and 7 for NPSLE discrimination. In the primary task-stratified analysis, pooled sensitivity was 0.91 (95% CI 0.86-0.94; 95% PI 0.56-0.99), and pooled specificity was 0.94 (95% CI 0.91-0.96; 95% PI 0.69-0.99), with low heterogeneity (I²=23.9% and 22.9%, respectively). DL models showed a sensitivity of 0.93 and specificity of 0.95, compared with 0.88 and 0.94 for traditional ML models. Certainty of evidence was high for most analyses but low for LN diagnosis because of inconsistency and imprecision. All studies were retrospective, and only 9 of 29 (31%) performed independent external validation. Overall risk of bias was high or unclear in 22 of 29 (75.9%) studies. No study reported model calibration, decision-curve analysis, or net clinical benefit. CONCLUSIONS: ML models showed promising diagnostic accuracy across 3 distinct SLE-related tasks, but wide PIs, limited external validation, and pervasive risk of bias restrict conclusions about real-world generalizability. Prospective multicenter studies with standardized tasks and reference standards, independent external validation, and formal assessment of calibration and clinical utility are required before clinical implementation.

Humans↗

PubMed2026

From fear to empowerment: the impact of employees AI awareness on workplace well-being - a new insight from the JD-R model.

PURPOSE: The primary purpose of the study was to explore the impact of health workers' awareness of artificial intelligence (AI) on their workplace well-being, addressing a critical gap in the literature. By examining this relationship through the lens of the Job demands-resources (JD-R) model, the study aimed to provide insights into how health workers' perceptions of AI integration in their jobs and careers could influence their informal learning behaviour and, consequently, their overall well-being in the workplace. The study's findings could inform strategies for supporting healthcare workers during technological transformations. DESIGN/METHODOLOGY/APPROACH: The study employed a quantitative research design using a survey methodology to collect data from 420 health workers across 10 hospitals in Ghana that have adopted AI technologies. The study was analysed using OLS and structural equation modelling. FINDINGS: The study findings revealed that health workers' AI awareness positively impacts their informal learning behaviour at the workplace. Again, informal learning behaviour positively impacts health workers' workplace well-being. Moreover, informal learning behaviour mediates the relationship between health workers' AI awareness and workplace wellbeing. Furthermore, employee learning orientation was found to strengthen the effect of AI awareness on informal learning behaviour. RESEARCH LIMITATIONS/IMPLICATIONS: While the study provides valuable insights, it is important to acknowledge its limitations. The study was conducted in a specific context (Ghanaian hospitals adopting AI), which may limit the generalizability of the findings to other healthcare settings or industries. Self-reported data from the questionnaires may be subject to response biases, and the study did not account for potential confounding factors that could influence the relationships between the variables. PRACTICAL IMPLICATIONS: The study offers practical implications for healthcare organizations navigating the digital transformation era. By understanding the positive impact of health workers' AI awareness on their informal learning behaviour and well-being, organizations can prioritize initiatives that foster a learning-oriented culture and provide opportunities for informal learning. This could include implementing mentorship programs, encouraging knowledge-sharing among employees and offering training and development resources to help workers adapt to AI-driven changes. Additionally, the findings highlight the importance of promoting employee learning orientation, which can enhance the effectiveness of such initiatives. ORIGINALITY/VALUE: The study contributes to the existing literature by addressing a relatively unexplored area - the impact of AI awareness on healthcare workers' well-being. While previous research has focused on the potential job displacement effects of AI, this study takes a unique perspective by examining how health workers' perceptions of AI integration can shape their informal learning behaviour and, subsequently, their workplace well-being. By drawing on the JD-R model and incorporating employee learning orientation as a moderator, the study offers a novel theoretical framework for understanding the implications of AI adoption in healthcare organizations.

Humans↗