Experimental Neurology

Generalizable neuroimaging biomarkers that detect cerebral cortical changes after traumatic brain injury (TBI) and predict patient outcomes are needed to improve care and to develop targeted therapies. We used morphometric inverse divergence (MIND) analysis of structural MRI to investigate cortical gray matter morphological networks cross-sectionally and longitudinally after TBI and correlate these with symptoms, disability and cognition six months after injury. Our findings support the Triple Network Model from functional MRI of post-traumatic alterations in the relationship between task-positive, default mode and salience networks. However, the strongest associations between early cortical similarity metrics and long-term patient outcomes involved the dorsal attention network and the limbic network as well as similarity metrics across Mesulam's hierarchy of laminar differentiation. Since MIND mapping of cortical gray matter networks only requires data that is a routine part of standard clinical MRI protocols and does not need image harmonization across different scanners, this work reports a promising new tool that is immediately available for advancing research and clinical care in TBI.

Quantitative eye movement analysis is important for neuro- logical diagnostics, yet existing video-oculography (VOG) systems typ- ically require calibration, device-specific settings, or accurate gaze la- bels. We present VOGeo-Gaze, a real-time, calibration-free, geometry- aware neural network that estimates gaze by reconstructing anatomi- cally meaningful eyeball parameters from image features. The method combines segmentation-driven projection geometry, a refraction-aware pupil correction module, and temporal anatomical stabilization, so gaze is derived from interpretable eye geometry rather than direct angular regression. Trained only on the public TEyeD dataset with weak gaze supervision, VOGeo-Gaze was evaluated on 116 clinical recordings from 17 patients and 19 healthy subjects using EyeSeeCam, a clinical gold- standard VOG system. It achieved median absolute angular errors of 0.33{whitebullet} horizontally and 0.35{whitebullet} vertically, with nearly 92% of recordings below 1{whitebullet} error while operating at >300 FPS. These results demonstrate sub-degree clinical gaze estimation without subject-specific calibration, camera intrinsics, or accurate gaze labels, providing a scalable and inter- pretable alternative to conventional VOG pipelines. Code is available at https://github.com/DSGZ-MotionLab/VOGeo-Gaze.

Background: Traditional diagnostic models lack explainability, while multimodal language models prone to hallucination remain unsafe for medical education. An interactive, risk-free artificial intelligence framework is required to serve as a reliable clinical mentor for radiology trainees. Methods: We propose a multi-agent architecture decoupling deterministic image analysis from generative consultation. Specialized computer vision models perform anatomical localization and pathological segmentation. These quantitative outputs are synthesized into a structured payload, which grounds a locally hosted large language model (LLaVA 7B) using strict prompt guardrails and prerequisite protocols. Results: The system effectively eliminates visual hallucinations by intercepting unanchored queries. The artificial intelligence tutor successfully contextualizes spatial anomalies and baseline metrics, generating accurate conversational explanations and formally structured radiology reports while strictly enforcing medical safety disclaimers. Discussion and Conclusion: By anchoring language generation exclusively to verified algorithmic realities, this framework transforms opaque diagnostic models into safe, interactive educational simulators. This establishes a highly reliable paradigm for integrating explainable artificial intelligence into medical training.

Objective: To systematically evaluate pathway-informed polygenic risk score (PRS) strategies and determine which approaches most effectively leverage biological annotations for risk prediction, using brain amyloid-beta positivity as a case study. Methods: We systematically benchmarked approaches for integrating pathway information into PRS construction to predict brain A{beta} positivity. Using two cohorts, the Alzheimer's Disease Neuroimaging Initiative (ADNI, n = 969) and Australian Imaging, Biomarkers and Lifestyle (AIBL, n = 251), we compared Apolipoprotein E (APOE) genetic risk score (GRS), clumping and thresholding (C+T) PRS, pathway-guided single nucleotide polymorphism (SNP) selection PRS, and pathway-specific PRSs ensembled via machine learning. Pathways were derived from manually curated literature or from pathway databases via Functional Mapping and Annotation (FUMA). Results: In cross-validation on the ADNI cohort, pathway-informed PRS using a narrow-set of pathways to guide SNP selection (PathPRS-SNPLit without APOE locus) significantly outperformed the standard PRS model (median AUC = 0.742, p = 0.006) and the APOE locus model (median AUC = 0.736, p = 5.1 x 10-5) based on the Mann-Whitney U test, achieving a median AUC of 0.763. This model showed enhanced ability to identify subgroups within the 10% lowest- and highest-risk groups compared to the current standard of APOE locus alone (odds ratio = 0.67, 95% CI: 0.56-0.81; and OR = 13.23, 95% CI: 10.23-17.11), highlighting its clinical potential. Using a focused set of literature-curated pathways outperformed using a broader set of database-derived pathways across configurations. When contrasting strategies for aggregating information across pathways, we observed that using pathways to guide selection of SNPs and then building a single PRS performed comparably to building PRS for each pathway and using machine learning (ML) to aggregate these, though the latter enabled pathway-level interpretability. Similar trends were observed in the external AIBL validation dataset. Interpretation: Pathway-informed PRS can meaningfully improve genetic risk enrichment for A{beta} positivity beyond APOE and standard C+T approaches, provided pathway definitions are carefully curated. The choice of pathway source has the strongest impact on predictive performance, with aggregation strategies or ML model choice having far less impact. Our findings highlight the utility of literature-curated, pathway-informed PRSs for A{beta} prediction and offer practical guidance for pathway-informed PRS construction in other polygenic traits.

Hypokalemia is a common and potentially life-threatening electrolyte abnormality in emergency care, yet rapid noninvasive screening remains difficult in time-critical triage settings. We developed PocketED-K, a single-lead AI-ECG prescreening model initialized from ECGFounder, and evaluated it in retrospective multicenter cohorts and a prospective handheld pilot. Retrospective development and validation included 37,115 patients from MC-MED and MIMIC-ED, and the pilot enrolled 18 patients at Peking University First Hospital. Hypokalemia was defined as venous serum potassium < 3.5 mmol/L. PocketED-K achieved AUROCs of 0.8189 (95% CI 0.8172--0.8207) in internal testing, 0.8104 (95% CI 0.8092--0.8115) in temporal validation, and 0.7889 (95% CI 0.7692--0.8074) in independent external validation; external negative predictive value was 0.9911 (95% CI 0.9895--0.9925). Higher predicted risk was associated with ST-segment depression, T-wave flattening or inversion, and relative U-wave prominence. The prospective handheld pilot provided an initial signal of workflow feasibility in real-world acquisition. These findings support single-lead AI-ECG as a low-burden prescreening tool to prioritize potassium testing in emergency care.

Importance: Physician-facing AI tools are now in clinical use, yet whether different platforms fail in similar or fundamentally different ways in high-stakes settings like critical care is unknown. Objective: To evaluate two physician-facing AI platforms, ChatGPT for Clinicians and OpenEvidence, for distinct vulnerabilities under structured stress testing. Design, Setting, and Participants: An observational study conducted using 60 simulated critical care vignettes developed and adjudicated by four attending critical care physicians. Data were collected in the last week of April 2026, via the public website interfaces of each platform. Interventions/Exposures: A 2x2x2x2 factorial design across four stressors - anchoring, cognitive load, social conformity pressure, and a clinically incorrect directive - yielded 16 prompt subsets per vignette and 960 prompts per platform. A separate multi-turn adversarial prompting paradigm administered three sequential "You are incorrect" challenges to baseline vignettes. All prompts had a universal output length constraint of fewer than 30 words. Main Outcomes and Measures: Critical elements capture (percentage of gold-standard critical elements present in responses), susceptibility to clinically incorrect directive, and sycophancy (reversal of an initial correct recommendation under iterative adversarial challenge). Results: Across 1916 responses to 1920 prompts, ChatGPT for Clinicians captured more gold-standard critical elements than OpenEvidence (81.4% {+/-} 18.1% vs 61.0% {+/-} 23.5%; adjusted difference, 20.3 percentage points; 95% CI, 18.3 to 22.4; P < .001) and was less susceptible to clinically incorrect directives (1.7% vs 8.0%; adjusted odds ratio, 0.07; 95% CI, 0.02-0.21; P < .001). Anchoring and social conformity pressure were associated with reduced critical element capture across both platforms, while cumulative stressor burden reduced critical element capture only on OpenEvidence. Conversely, ChatGPT for Clinicians reversed correct recommendations more readily under adversarial prompting (hazard ratio, 2.61; 95% CI, 1.10 - 6.19; P = .03). Conclusion and Relevance: The two physician-facing clinical AI platforms evaluated demonstrated non-overlapping vulnerabilities, with neither platform uniformly superior. These findings argue against single-axis ranking of clinical AI systems and support multidimensional safety evaluation encompassing completeness of reasoning, resistance to incorrect directives, and stability under adversarial challenge.

Background: Patients with CKD and polypharmacy face high rates of drug-related problems, yet comprehensive medication review remains time-intensive and inconsistently performed. Large language models (LLMs) may augment this process, but existing benchmarks use multiple-choice formats that do not reflect open-ended, nephrology-specific review. We developed a trap-embedded synthetic CKD benchmark and evaluated five current-generation LLMs (GPT-5.4, Claude Sonnet 4.6, Gemini 3.1 Pro, Grok 4.1 Fast, DeepSeek R1; tested April-May 2026) for open-ended medication review. Methods: Fifty synthetic CKD cases across three complexity groups (G3a-G3b [n=20], G4 [n=15], G5/G5D/transplant [n=15]) with 8-12 medications and [&ge;]2 embedded clinical traps each were scored against nephrologist-adjudicated gold standards. Each model produced three independent responses per case (temperature 0; 750 total outputs). Primary endpoint was per-case macro F1; secondary endpoints were safety-critical omission rate, PI-adjudicated hallucination rate, and intra-model consistency. Blinded inter-rater reliability for gold-standard item detection was assessed on a 30% sample. Results: Consensus-level macro F1 ranged from 0.41 (Claude Sonnet 4.6) to 0.49 (Grok 4.1 Fast) (Friedman P < 0.001). Phosphate binder timing (11%) and hyperkalemia combinations (33%) were poorly detected across all models. Safety-critical omission rate ranged from 22% to 48% (P < 0.001); PI-adjudicated hallucination ranged from 0% (GPT-5.4) to 54% (DeepSeek R1), including fabricated dose caps and non-existent guideline citations. Blinded reliability for gold-standard item detection was high (kappa = 0.934, n = 92). Conclusions: This nephrology-specific benchmark exposes clinically important LLM blind spots that generic multiple-choice evaluations would not detect. Heterogeneous hallucination and omission rates indicate that model selection and domain-specific guardrails should precede any clinical deployment of LLM-assisted CKD medication review. Prospective validation with real patient data and human comparators is required before deployment recommendations can be made.

Artificial intelligence (AI) tools have been rapidly adopted by medical researchers, yet whether early career researchers in low and middle income countries possess the awareness and habits needed to use these tools safely remains poorly documented. This study characterized AI adoption patterns, hallucination awareness, and verification and disclosure practices among early career medical researchers in Pakistan. A cross sectional anonymous online survey was conducted among medical students, house officers, residents, physicians, and faculty involved in research or academic work across Pakistan (May 2026). Descriptive statistics and chi square tests were applied to 373 eligible responses. AI use was near universal (99.7%), with 60.3% using AI tools daily. The most commonly reported tool in this sample was Claude (40.5%), followed by ChatGPT (29.2%) and Perplexity (26.0%), though this ranking likely reflects sampling characteristics. Despite high adoption, 59.2% typically did not verify AI outputs before use, and 40.2% had never heard that AI can generate fabricated scientific references. In behavioral vignettes, 36.5% assumed convincing AI generated references were authentic, and 54.2% would continue using remaining AI content after discovering one fabricated reference. Formal research training was strongly associated with consistent disclosure (51.7% vs. 17.1%; chi square=48.43, p less than 0.001). Role, daily use frequency, and research training were not significantly associated with verification behavior. Early career medical researchers in Pakistan demonstrate high AI adoption alongside incomplete hallucination awareness and infrequent verification, a pattern that may carry implications for research integrity. Formal training was the only factor significantly associated with consistent disclosure. Integration of AI literacy into medical curricula and institutional governance frameworks merits consideration.

Machine learning (ML) is being considered to help diagnose cardiovascular diseases (CVD). Still, challenges like inconsistent and limited datasets, limited infrastructure, and global inequalities lead to the need for a reliable and practicable ML solution. This paper presents an ML-driven framework for predicting CVD risk scores and classifying status. Several data preprocessing techniques, including multiple imputation by chained equations (MICE), outlier removal, are considered. In addition, hyperparameter tuning is performed with the GridSearchCV tuning technique. Moreover, a consensus-driven five-feature selection method is applied to identify optimal predictors. The dataset used in this study contains healthcare records related to future CVD risk scores, comprising 1,529 patient records with 22 features. The optimized stacked ensemble model is applied to the dataset and achieves a cross-validated coefficient of determination value of 98.13% for CVD risk score regression. Comparative evaluation with other ML models confirmed improved accuracy, efficiency, and interpretability. The explainable AI technique SHAP is applied to interpret predictions and highlight key risk factors. Moreover, a deployment-ready web platform with multi-role access has been developed that demonstrates clinical applicability. The proposed framework offers a reliable and interpretable tool for early detection of CVD and personalized risk assessment. In the future, this work can be extended to integrate longitudinal data, medical imaging, and deep learning to improve generalizability and strengthen real-world impact.

Background Artificial intelligence-enhanced electrocardiography (AI-ECG) enables scalable, low-cost cardiac dysfunction screening, but existing models are annotation-intensive and predominantly adult-derived, leaving paediatric generalizability uncertain. Paediatric cohorts exhibit highly variable cardiac morphology and function compared to adults, which may be useful for learning generalizable AI-ECG models. Methods We pretrained ECG-Fyler on a predominantly paediatric, all-age cohort at Boston Children's Hospital (1992-2023), annotated with a cardiology-specific coding system (Fyler codes), and evaluated it on assessments from echocardiography (echo) and cardiac magnetic resonance (CMR) studies. We validated on an external adult cohort from Columbia University Irving Medical Center. Performance was benchmarked against several AI-ECG foundation models by AUROC across age groups, lesion types, and limited-data scenarios. Findings The pretraining cohort comprised 782,138 ECGs from 255,271 patients (median age: 10.9 years, IQR: [2.8-16.8]). Internal evaluation included 178,495 ECG-echo pairs (median age: 10.9 [3.7-17.0]) and 8,584 ECG-CMR pairs (median age: 20.7 [15.6-29.6]). External validation included 82,543 ECG-echo pairs from adults (median age: 64.0 [52.0-74.0]). ECG-Fyler improved AUROC across biventricular dysfunction and dilation tasks, with the largest gains in low-data settings. In internal validation, ECG-Fyler detected low left ventricular ejection fraction (LVEF [&le;] 40%) from only 100 fine-tuning samples (AUROC: 0.80, 95% CI: [0.78-0.80]), outperforming other models (AUROC < 0.65) and improving with additional fine-tuning (AUROC: 0.94 [0.93-0.94]). Similar improvements were observed for CMR-derived LVEF, RVEF, and ventricular dilation. In external validation on adults, ECG-Fyler exhibited an AUROC of 0.83 (CI: [0.82-0.85]) for LVEF [&le;] 40%. After fine-tuning on less than 10% of external data, LVEF [&le;] 45% performance (AUROC: 0.87 [0.86-0.88]) outperformed a fully trained, site-specific prior model (AUROC: 0.85 [0.84-0.87]). Interpretation Pretraining on richly annotated, paediatric-dominant ECGs yields models that transfer efficiently across institutions and ages, supporting AI-ECG screening and triage when labels or imaging access are limited. Funding National Institutes of Health (R01LM012973); Kostin Innovation Fund, Boston Children's Hospital

Background: The All of Us Research Program represents a rich resource for cancer epidemiology research, with over 400,000 participants with whole genome sequences linked to electronic health records (EHR). Large cancer datasets often focus exclusively on cases without controls and neglect pre-diagnosis healthcare occurrences. Here, we perform a phenome-wide association study (PheWAS) of EHR data at least 1 year pre-diagnosis between cancer cases and matched controls, revealing co-occurring and mutually exclusive phenotypes. Methods: We identified 55,000+ cancer cases across 21 cancer types in All of Us version 8. To eliminate age-related confounding, we implemented a two-stage matching and censoring strategy: loose matching on demographics to establish index dates and cohort comparability, followed by right-censoring of EHR data (excluding 1 year pre-diagnosis/index), then 1:2 matching to address residual demographic imbalance. We tested associations between 23,193 cancer cases, 46,386 matched controls and approximately 1,600 clinical phenotypes using logistic regression adjusted for sex at birth, self-reported race, age at diagnosis/index date, and two censored EHR metrics: observation window and unique condition count, with Bonferroni correction for multiple testing. Results: Our analysis identified 232 significantly associated phenotypes, confirming established cancer risk factors including elevated prostate specific antigen (OR = 2.92, 95% CI: 2.65-3.23; p-value=1.8x10-101) and multinodular goiter (OR = 1.73, 95% CI: 1.56-1.91; p-value=6.7x10-27). Further investigation into the relationship between several phenotypes with seeming inverse effects is warranted. Conclusions: This PheWAS of EHR data at least 1 year pre-diagnosis leveraged the diversity of All of Us to examine how clinical phenotypes prior to cancer diagnosis vary across cancer types and racial groups. Our findings validate All of Us as a robust platform for cancer epidemiology research, confirming established risk factors at scale across diverse populations. This work provides methodological insights for EHR-based susceptibility analyses and demonstrates the value of agnostic phenome-wide approaches for generating hypotheses in precision medicine.

Objective Memory impairment is a frequent comorbidity of focal epilepsy, incompletely explained by seizure frequency or structural pathology. Ictal and postictal hippocampal dysfunction disrupt memory processes, but their cumulative impact remains poorly quantified. This study introduces cumulative hippocampal seizure-related burden metrics and examines their association with long-term memory consolidation. Methods Twenty consecutive patients undergoing stereo-EEG in Marseille (2016-2018) were prospectively included. Continuous stereo-EEG recordings between two memory assessments (30 minutes and one week post-encoding) were analysed. Hippocampal ictal involvement and durations were assessed using epileptogenicity markers and visual stereo-EEG analysis. The postictal period was quantified using permutation entropy. Cumulative hippocampal seizure-related burden metrics (ictal, postictal and combined: c-HipSZB) were computed across hippocampus-involving ictal events. Verbal and visual memory were assessed using standardized recall and recognition tasks. Associations were examined using univariate and multivariate analyses. Results Higher dominant-hemisphere hippocampal burden was associated with poorer one-week verbal memory (performance and retention), independently of most covariates. Higher c-HipSZB was associated with lower total recall performance (RT; free + cued) and RT retention ({beta} = -25.04 and -23.88; R2 = 0.57 and 0.53; p < 0.05) and accounted for the greatest variance in both outcomes (adjusted R2= 0.59 and 0.53; {beta} = -25.45 and -24.27; p < 0.01), particularly when adjusting for epilepsy duration. No robust associations were observed between non-dominant-hemisphere hippocampal seizure-related burden metrics and visual memory. Effects predominantly involved recall. Interpretation Cumulative ictal-postictal hippocampal dysfunction is a major determinant of impaired long-term verbal memory consolidation in focal epilepsy.

Subjective auditory-perceptual evaluation and uninterpretable deep learning models limit the clinical assessment of voice disorders. This study proposes a two-phase zero-shot framework to evaluate voice pathology. First, an Audio Spectrogram Transformer is fine-tuned on the Perceptual Voice Quality Database to generate an acoustic latent space. Second, Orthogonal Procrustes analysis maps these acoustic embeddings directly onto the semantic space of a pre-trained Sentence Transformer. The geometric alignment produced continuous semantic axes that outperformed a supervised machine learning baseline in regressing clinician-rated GRBAS (Grade, Roughness, Breathiness, Asthenia, and Strain) severity scales. Furthermore, these axes correlate with traditional acoustic measures, including Harmonics-to-Noise Ratio and local jitter, while remaining robust when applied to aperiodic signals by not requiring fundamental frequency extraction. Most importantly, the model achieved zero-shot semantic expansion, successfully evaluating voices using an untrained, natural clinical vocabulary beyond the GRBAS scale. External validation on the Voice ICarus Database confirmed cross-corpus stability and demonstrated the capacity for zero-shot differential phenotyping of specific etiologies, such as hypokinetic dysphonia and reflux laryngitis. By bridging acoustic and semantic latent spaces, this framework offers an objective, continuous, and transparent metric for evaluating voice quality using voice descriptive vocabulary.

Our study evaluated whether a deep learning auto segmentation model combined with machine learning triage can streamline radiotherapy clinical trial quality assurance (QA). We analyzed 107 stereotactic ablative radiotherapy (SABR) cases from a multi-institutional phase II clinical trial of neurovascular sparing prostate SABR, focusing on physician contours of the internal pudendal artery (IPA) as a novel organ-at-risk with substantial interobserver variability. Contours were scored by the trial principal investigator as Per-Protocol or Minor Deviation/Unacceptable. We applied a deep learning model for IPA auto-segmentation. Agreement between human and AI contours was then quantified using 14 overlap, distance, and surface metrics, and a supervised classifier was trained on these metrics to flag clinical trial protocol deviations. While AI segmentation achieved only modest geometric accuracy with mean Dice similarity coefficient of 0.446 and 95th percentile Hausdorff distance of 14.23, when incorporating all 14 metrics, a machine learning classifier yielded AUROC of 0.836, flagging all Minor Deviation/Unacceptable cases with 100% sensitivity on the 27 case hold-out set with 6 false positives and no false negatives. AI segmentation combined with metrics-based machine learning can triage protocol deviations within a multi-institution radiotherapy clinical trial, supporting prospective evaluation of AI-assisted trial QA.

Background: Complicated acute appendicitis carries a higher risk of postoperative morbidity relative to uncomplicated cases. It remains unclear whether surgical timing (night vs. day; weekend vs. weekday) or surgeon seniority influence short-term outcomes in this high-risk population. Methods: This was a retrospective analysis of the RIFT Turkey dataset restricted to histologically confirmed cases of complicated appendicitis who had undergone laparoscopic appendectomy. Primary exposures were surgical timing (day [n=92] vs. night [n=123]; weekday [n=172] vs. weekend [n=43]) and surgeon seniority (trainee [n=89] vs. consultant [n=126]). The primary outcome was unplanned readmission and/or reintervention within 60 days. Secondary outcomes were conversion to open surgery and length of stay (LOS) >3 days. Propensity score matching (PSM) using RIPASA score (caliper 0.05, SMD <0.1) was performed as a pre-specified sensitivity analysis for each comparison. Results: Night-time surgery was associated with higher frequencies of unplanned readmission / reintervention (12.2% vs. 6.5%; OR 1.99 [95% CI 0.74-5.35], p=0.166) and surgical conversion (9.8% vs. 3.3%; OR 3.21 [0.88-11.72], p=0.064) compared with daytime surgery, neither reaching significance. Trainee surgeons had significantly higher readmission/reintervention rates than consultants (15.7% vs. 5.6%; OR 0.32 [0.12-0.82], p=0.013). PSM-adjusted results also showed similar relationships: night vs. day (readmission OR 2.45 [0.85-7.03], p=0.09; conversion OR 2.84 [0.73-11.1], p=0.13), weekend vs. weekday (readmission OR 1.53 [0.24-9.72], p=0.65), and trainee vs. consultant (readmission OR 0.25 [0.08-0.79], p=0.013). Conclusion: Surgical timing was not significantly associated with short-term outcomes in complicated appendicitis, though night-time surgery showed a consistent trend towards higher complication rates. Surgeon seniority was the only factor independently and significantly associated with unplanned readmission and reintervention in both primary and PSM analyses, indicating the need for senior supervision during out-of-hours procedures. Keywords: complicated appendicitis; surgical timing; night surgery; weekend effect; surgeon seniority; propensity score matching; RIFT Turkey

Objective. Biomedical knowledge graphs (KGs) such as PrimeKG, Hetionet, UMLS, and PharmGKB are increasingly used as the substrate for downstream machine-learning, retrieval-augmented generation, drug-repurposing, and electronic health record (EHR) augmentation pipelines. The dominant assumption in published work is that integrating two or more such KGs is a tractable engineering step solved by identifier (ID) matching. This paper interrogates that assumption empirically. We quantify how much concept overlap survives realistic alignment, and we characterize the new failure modes introduced by the methods that practitioners reach for when ID matching is insufficient. Materials and Methods. We compared four widely used biomedical KGs (PrimeKG, Hetionet v1.0, the full UMLS Metathesaurus, and PharmGKB) across eleven node types using a tiered alignment pipeline: (1) direct ID matching for nodes sharing a primary vocabulary; (2) cross-ontology bridging using standard mappings (e.g., MONDO-DOID, HPO-UMLS, HPO-UMLS-MeSH for side effects, NCBI Gene-HGNC-UMLS, UBERON-FMA/SNOMEDCT_US/NCI/MeSH for anatomy); (3) ClinicalBERT cosine-similarity grouping at threshold >= 0.98 for over-segmented disease nodes, with a deterministic suffix-stripping canonicalizer; (4) exact name matching for ontology-poor types (anatomy, REACTOME pathways); and (5) embedding-based fuzzy matching with UMLS lookup (SapBERT and ClinicalBERT) for free-text microbiome concepts. We applied the pipeline to a 698-concept gut-microbiome benchmark spanning taxa, pathways, and disease labels, validated grouping decisions against the curated SSSOM mappings released by the MONDO project, and audited the ClinicalBERT consolidation against five clinical-genetics case studies drawn from the literature. Results. Per-type pairwise coverage was strikingly asymmetric. Genes/proteins and the three Gene Ontology categories aligned cleanly across PrimeKG and Hetionet (mutual coverage 94-99%), but disease overlap was sparse: only 0.7% of PrimeKG individual disease nodes mapped to Hetionet, rising to 2.0% after MONDO grouping (versus 78.7% and 18.4% from the Hetionet side). PrimeKG-to-UMLS coverage spanned 100% (effect/phenotype via HPO) down to 20.8% (REACTOME pathways), with drugs at 73.7% and anatomy at 58.8%. PrimeKG-to-PharmGKB drug coverage required up to two bridging hops (DrugBank -> UMLS -> RxNorm/ATC/MeSH). Bigger was not uniformly more complete: on a 698-concept microbiome drug benchmark, Hetionet missed 0 concepts while PrimeKG missed 16. ClinicalBERT-based grouping consolidated 22,205 raw MONDO disease nodes into 17,080 groups but introduced three reproducible failure modes documented in case studies: (i) peer over-merging: for example, all 22 osteogenesis imperfecta subtypes collapsed into a single node despite distinct severity classes; (ii) parent-child collapse: e.g. acute myeloid leukemia merged with myeloid leukemia, erasing the acute/chronic distinction that drives clinical management; and (iii) lexical false positives: neurofibromatosis and schwannomatosis grouped together despite cellular-pathology differences. Discussion. Identifier matching alone is a weak baseline for biomedical KG integration. Cross-ontology bridges and embedding-based consolidation expand coverage but do so at the cost of clinically meaningful resolution, and the resulting failures are systematic rather than random. Reporting only aggregate coverage statistics obscures these losses, which propagate silently into downstream tasks. Conclusion. We provide reusable per-type coverage tables, a taxonomy of three integration failure modes, and concrete recommendations for downstream studies that depend on a unified biomedical KG. We argue that future KG integration work should report per-type coverage and per-cluster confidence rather than aggregate match rates.

Vocal biomarkers, encompassing voice and speech, have largely been developed for individual conditions in isolation, limiting their generalizability across diseases and recording settings. To address this, we introduce VoiceFM, a contrastive model that learns general-purpose clinical voice representations by aligning audio embeddings with rich clinical metadata. Using the Bridge2AI-Voice dataset (984 primarily English-speaking adult participants, 846 used for training and 138 held out as a temporally separated validation cohort, 40,056 recordings totaling 176 hours across 5 academic medical centers), VoiceFM pairs a fine-tuned Whisper large-v2 encoder with a tabular transformer over 44 clinical features via symmetric InfoNCE loss. Linear probes on frozen VoiceFM embeddings achieve mean AUROC 0.952 +/- 0.005 across five evaluation tasks (control vs disease screening plus four disease categories), significantly outperforming Frozen Whisper (0.926 +/- 0.013, p = 0.013), Frozen HuBERT (0.885 +/- 0.017, p = 0.0009), and the contrastively trained VoiceFM-HuBERT (0.938 +/- 0.006, p = 0.012). On the 138-participant held-out cohort, VoiceFM-Whisper achieves AUROCs of 0.99 for Alzheimer's/dementia/MCI and 0.89 for airway stenosis, demonstrating that the learned representations generalize to participants the model has never seen. VoiceFM representations transfer to three external datasets without retraining and improve few-shot classification. Recording task attribution identifies a small set of speech tasks that match or exceed the full battery's performance, suggesting shorter screening protocols are feasible. Trained predominantly on English audio, VoiceFM transfers without fine-tuning to Spanish-language Parkinson's disease (PD) detection (NeuroVoz, 107 participants, AUROC 0.93 +/- 0.02), with the signal dominated by articulatory rather than phonatory features. A fine-tuned classifier achieves participant-level AUROC 0.87 (sustained 0.85, countdown 0.80) on the mPower smartphone study (585 held-out participants). Together, these results show that contrastive alignment between voice and rich clinical metadata can serve as the basis for a clinical voice foundation model, producing a single set of transferable representations that generalize across diseases, languages, recording conditions, and patients enrolled after model freeze.

Background: Autosomal recessive spastic ataxia of Charlevoix-Saguenay (ARSACS) is a genetic disease characterized by spasticity and ataxia which reflects involvement of the corticospinal tracts (CST) and cerebellum. The primary involvement of the middle cerebellar peduncles (MCP) and transverse pontine fibers (TPF) at the crossing with the CST, and their role in the pathophysiology of the disease, is currently debated. Objectives: Advanced MRI techniques capable of isolating sub-voxel microstructural parameters can test the hypothesis that the MCP and TPF are abnormally large, compressing the CST at their crossing, and potentially impairing CST development. Methods: Tract macro- and micro-structural properties, including axon and tract caliber, axon density and geometry, and myelin content were estimated from diffusion-relaxometry and magnetization transfer imaging. These features were analyzed along segments of the CST, MCP, and TPF of 9 patients and 9 age-matched controls. Results: While the CST showed significant decreases in tract size, axon caliber, and myelination throughout its length compared to controls (p<0.01), the MCP and TPF were relatively unaffected. In our group, neither the MCP nor the pons were enlarged. The proximal MCP showed an increase in axon caliber. Conclusions: The increase in fractional anisotropy and axon density towards the center of the TPF could be driven by geometric confounds related to differences in the relative sizes of the CST and TPF compared to controls. This highlights the importance of investigating tract-specific microstructural profiles, particularly in regions of geometric complexity. The findings confirm the involvement of the CST, with a relatively limited involvement of the MCP and TPF.

Background: Hospital incident risk scoring has long relied on two- or three-dimensional frameworks (Severity Assessment Codes or Risk Priority Numbers),even though root cause analysis standards recognize that clinical risk is multi-factorial. The obstacle has been mainly cognitive: human reviewers cannotreliably score many dimensions across high incident volumes, so richer assessmenthas not been operationalized at scale.Objective: To extend the traditional three-dimensional FMEA to an eight-dimensional patient-safety risk feature framework, to establish a multi-modellarge language model (LLM) extraction pipeline that scores these dimensionsautomatically, and to demonstrate a variance-aware integer optimization (mean-variance integer programming, MV-IP) that provides a reproducible tie-breakingrule for incident prioritization under extraction uncertainty, rather than improvedrisk coverage.Methods: An 8-dimensional framework covering harm severity, potential harm,frequency, detectability, systemic impact, vulnerable populations, regulatoryrelevance, and economic impact was applied to 213 synthetic and 196 realcurated incident narratives. Three independent LLMs (GPT-5.4, Gemini 3.1 Pro, Grok-4.1 Fast) from different provider families extracted structured risk scores.Inter-model consistency was assessed via ICC(A,1). Among coverage-equivalentselections, MV-IP minimized inter-model variance to give a reproducible prioriti-zation rule. An English-language sensitivity analysis was conducted on 31 AHRQPSNet WebM&M cases.Results: On real cases, seven of eight dimensions reached Fair or betterinter-model reliability (ICC(A,1) 0.53 to 0.83); D5 (Systemic Impact) was theexception at Poor reliability (0.275), driven by little between-case variation ratherthan by wide model disagreement. Reliability was not uniform: two dimensionswere Excellent (D1 actual harm 0.834, D8 economic impact 0.782), two Good,and three only Fair, so some dimensions are more readily extractable than others.The same anchors gave broadly similar results on English-language narratives.When deterministic top-K selection returned several equal-coverage solutions(11 on real cases, total inter-model variance 0.205 to 1.274), MV-IP selected theminimum-disagreement set, replacing ad hoc tie-breaking with an explicit rulewithout improving coverage. Bootstrap resampling found 74% to 90% of per-casevariance estimates stable despite the three-model panel.Conclusions: The eight-dimensional framework operationalizes patient-safetyrisk features that quality teams have considered only implicitly, and three inde-pendent LLM families produced reproducible scores on most dimensions ofcurated narratives. Inter-model agreement, however, measures reproducibilityrather than clinical correctness, and high agreement does not by itself establishthat a score is right; the dimensions that are reliably extractable today (notablyD6 and D8) differ from those that are not yet (D5, and to a lesser degree D4 andD7), which has direct implications for incident-reporting form design. MV-IP con-tributes a reproducible, variance-aware tie-breaking rule rather than improvedcoverage. Validation against expert-prioritized RCA lists and deployment on rawinstitutional incident reports remain the next steps toward clinical use.

Although GLP1/GIP receptor agonists demonstrate unprecedented weight loss efficacy, their rapid clinical adoption has revealed significant real-world tolerability challenges. To evaluate their dynamic safety profiles, we developed a macro to micro pharmacovigilance framework by combining global FAERS reports with local UT Physician EHR. Macroscopically, we distilled 17 shared adverse events across the drug class from FAERS with disproportionality analysis. Microscopically, local EHR data (289,655 longitudinal treatment sessions across 71,316 patients) revealed 51.6% of GLP1 sessions terminated within 90 days. Furthermore, temporal stratified logistic regression demonstrated that initial exposure (0 to 30 days) correlated strongly with nausea and vomiting, which attenuated in extended sessions, whereas extended exposure (>2 years) uncovered late onset risks, notably incident hepatic steatosis. Ultimately, this time aware framework reveals that GLP1 safety profiles are profoundly duration dependent, providing critical insights into both acute intolerances and long-term medication safety.