Data in Brief

Introduction Air pollution is the largest environmental risk to human health. Children are disproportionately affected by air pollution and their exposure is amplified during physical activity. Observed concentrations of nitrogen dioxide in 1 in 4 London school playground exceeds the European limit, but the health impacts of air pollution exposure in London school playgrounds remain unexplored. Our study aims to assess and compare the acute changes in lung function and airway inflammation of primary school-aged children exercising in school playgrounds. Methods and analysis 330 children aged 8 to 11 years from ten London schools will be recruited to complete 90 minutes of physical activity and 90 minutes of rest in their school playground in a randomised crossover design. Pre-, post-, and 24-hour post-exposure oscillometry measurements will be performed with airway resistance at 5 Hz (R5) the primary physiological outcome. Nasal lavage samples will be collected pre-exposure and 24-hour post-exposure for analysis of inflammatory, oxidative, and vascular biomarkers, with IL-6 as the primary biological outcome. Mixed-effects regression models will examine associations between estimated pollutant exposures, exercise and physiological responses.

Background: Citrin deficiency, caused by biallelic pathogenic variants in SLC25A13, must be identified early to prevent serious complications such as hyperammonemia and liver failure. However, clinical diagnosis is often delayed due to its nonspecific presentation and limited sensitivity of amino acid-based newborn screening methods. Although genome-based evaluations are being investigated to address these issues, concerns about their cost, turnaround time, variant interpretation ability, and data handling highlight the need for a more practical yet reliable alternative. We investigated the feasibility of applying proteomic approach on dried blood spots (DBS), which are routinely used in newborn screening. Methods: We performed untargeted liquid chromatography-tandem mass spectrometry to analyze the proteome of DBS using a previously developed "non-targeted analysis of non-specifically DBS-absorbed proteins" (NANDA) workflow. SLC25A13 protein abundance was quantified in individuals with biallelic loss-of-function mutations, compound loss-of-function/missense mutations, and heterozygous carriers; this was also evaluated in healthy and diseased controls representing relevant differential diagnoses. To leverage proteomic information, we derived a multivariate proteomic signature using feature selection and evaluated its performance with leave-one-out cross-validation. Biological relevance was assessed by enrichment analysis, and complementary transcriptomics was performed using RNA sequencing. Results: A total of 7,474 proteins, including SLC25A13, were consistently detected in DBS. SLC25A13 was undetectable in individuals with biallelic loss-of-function mutations. However, individuals with compound loss-of-function/missense genotypes showed reduced but measurable SLC25A13 levels, comparable to those observed in heterozygous carriers. In contrast, a compact 15-protein signature accurately identified individuals with compound loss-of-function/missense genotypes (AUC, 0.99; sensitivity, 1.00; specificity, 0.95). The signature was enriched for Ca2+-response, and transcriptomics showed downregulation of genes related to multimodal ion channels in affected individuals compared to controls. Conclusions: DBS-based proteomic profiling may assist in the diagnosis of citrin deficiency through SLC25A13-quantification and a biologically plausible multivariate signature. More broadly, this strategy offers a promising new diagnostic layer for protein disorders, providing a proteomic readout in a clinically practical DBS format with potential utility for future diagnostic and screening applications.

One of the prominent methods in neuroimaging data processing is SSM-PCA, which is based on principal component analysis and allows for the identification of diagnostically significant patterns in the form of statistical maps. We developed software, PIE Toolbox, employs SSM-PCA and classification based on the obtained diagnostic patterns revealed from functional and structural tomographic brain imaging. The program supports the entire analysis pipeline including preprocessing of brain images, diagnostic patterns extraction, building classification models, and prediction based on them. The resulting diagnostic patterns are weighted principal components obtained through SSM-PCA, or their linear combinations. PIE Toolbox allows selection of relevant structural and functional brain patterns, computation of their expression values in regions of interest, classification using support vector machines, and evaluation of model performance via cross-validation. This approach enables the use of patterns as features of intergroup differences for individual diagnosis. The software has been validated on both simulated and ADNI datasets.

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.

Background and Objective: Access to real-world electronic health records (EHRs) remains limited by privacy, governance and annotation constraints, hindering the development of clinical natural language processing models. Realistic synthetic progress notes may provide EHR-like corpora that preserve clinically rigorous information on diagnoses, treatments, symptoms, imaging, laboratory findings and therapeutic trajectories without relying directly on sensitive patient records. This study evaluates whether large language models (LLMs) can generate realistic Spanish prostate cancer progress notes from published case reports, preserving clinical content, temporality and hospital-style conventions.

Background: Climate mitigation policies can lower air pollutant concentrations and deliver substantial health co-benefits. The French Ecological Transition Agency (ADEME) proposed four contrasting Transitions 2050 net-zero scenarios. We quantified mortality, morbidity, and health-economic co-benefits from projected PM2.5 and NO2 reductions across all four scenarios in continental France. Methods: Emission projections were input to the CHIMERE chemistry-transport model to estimate PM2.5 and NO2 concentrations for 2030 and 2050. Health impacts were assessed using disease-specific cessation-lag assumptions relative to 2019, covering premature mortality, morbidity, DALYs, and economic benefits across nine outcomes (hypertension, lung cancer, ischaemic heart disease, stroke, COPD, type-2 diabetes, acute lower respiratory infections, and asthma in children and adults). Findings: Population exposure is projected to decline by about 40% for PM2.5 and 70% for NO2 by 2050, with health gains remaining substantial and broadly equivalent across all four scenarios and modest differences between sufficiency-oriented and technology-driven pathways. Under delayed-impact assumptions, avoided premature deaths ranged from 21,300 to 22,100 for PM2.5 and 24,500 to 26,200 for NO2. Morbidity and disability-adjusted life year (DALY) reductions, as well as economic savings, spanned similarly; total avoided morbidity cases were 84,000-88,000, direct medical cost reductions were e1.0-1.1 billion/year, and intangible cost savings of e41-43 billion and e36-39 billion, respectively. Interpretation: Health co-benefits are substantial, consistent across contrasting scenarios, and increase markedly from 2030 to 2050. Explicitly incorporating these co-benefits into climate policy appraisals may strengthen the case for ambitious mitigation and improve decision-maker acceptability.

Sex work is diverse and includes a broad range of people and settings. Over the last thirty years, a large proportion of public health emergencies of international concern (PHEIC) have involved infections transmitted through sexual or close contact and in sexual networks (WHO 2024). Sex workers can face increased disadvantage in relation to these public health emergencies. Given the significant health inequalities sex workers can face, they should be eligible to receive targeted and tailored health support to reduce health protection risks (Hester 2019; Jeal and Salisbury 2004a). However, they are often not explicitly eligible for targeted and tailored support due to a lack of information on incidence, prevalence of disease, and even more basic data such as reliable estimates of the number of sex workers in the UK. Accordingly, the aim of this paper is to determine a population size estimate, with uncertainty, that is more robust than those currently available. In this study, we apply Bayesian Evidence Synthesis to bring together historic estimation efforts with recent ONS National Population Estimates and Genito-Urinary Medicine Clinics Attendance Data (GUMCAD) from the UK Health Security Agency (UKHSA). A key feature of our model is the embedding of uncertainty from each input study in model priors, hence propagating it through to our final estimate. The Bayesian evidence synthesis model estimated a total of 84,000 sex workers in the United Kingdom (95% credible interval: 49,000-130,000), representing 0.121% of the current UK population.

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.

Currently, the genetic architecture of Middle Eastern populations is underrepresented in global genomic databases. This gap increases the rate of Variants of Uncertain Significance (VUSs) and clinical misinterpretations of genomic data especially in Middle Eastern populations. Whole exome sequencing was conducted on 90 healthy individuals from Jordan and the data were analysed using Principal Component Analysis (PCA) and multi-computational filtering. PCA revealed a double ancestry (EUR-AFR) admixture rather than a triple admixture (EUR-AFR-AMR). More than 3,500 populations-specific variants (PSVs) were identified, of which 72% were singletons. Additionally, 19 variants were significantly enriched compared to the maximum allele frequencies in public global databases (Fisher's exact test with Benjamini-Hochberg false discovery rate correction, p-value < 0.05). Consequently, the results suggest the reclassification of variants of Uncertain Significance (VUS) which reside in the ECE2 gene to likely benign and the variants of Conflicting Classification of Pathogenicity in the genes IL1RN and THPO to benign based on the significant allele frequency (AF=0.0389, p-value < 0.05). Furthermore, a pathogenic ClinVar variant was identified in a healthy individual, warranting careful interpretation. The findings underscore the importance of identifying PSVs in order to minimize or even prevent clinical misdiagnosis and highlight the unique genetic signature in Jordan. The study serves as a foundational resource for precision medicine in the region.

We present the TopBrain 2025 Challenge, the first benchmark for fine-grained multiclass segmentation of the whole brain vasculature in both computed tomography angiography (CTA) and magnetic resonance angiography (MRA). Building on the TopCoW challenge, TopBrain scales vessel annotation from the Circle of Willis to the entire brain, introducing a dataset of 90 annotated volumes across 48 landmark vessel classes spanning arterial and venous systems, of which 50 training volumes are publicly released. Vessel definitions were consolidated from established neuroanatomical references into a unified annotation scheme, and vessel caliber measurements along the centerline are reported for the first time across the whole brain vascular anatomy. To address the unique challenges of multiclass brain vessel segmentation, we propose an evaluation framework that accounts for detection in segmentation performance, assesses anatomical plausibility, and introduces novel contamination metrics that characterize inter-class prediction errors. Fifteen teams from over 220 registered participants submitted algorithms to the benchmark. The top-performing teams built on nnUNet with principled system design choices, achieving around 80% Dice scores, near-zero invalid neighbor counts, over 60% F1 scores for side-road vessels, and below 18% foreground contamination ratio. Larger vessels are easier to segment, while smaller and more complex vessels remain the true bottleneck. The annotated datasets and podium-finish algorithms are made publicly available on Zenodo.

INTRODUCTION: Alzheimers disease and related dementias (AD/ADRD) arise from cumulative environmental, social, behavioral, and biological influences across the life course. The neural exposome framework conceptualizes how exogenous, behavioral, and endogenous factors interact to shape brain health; however, its application to preclinical AD/ADRD research, particularly in rural populations, remains limited. METHODS: We developed and piloted a community-embedded, decentralized research model to operationalize the neural exposome framework among cognitively unimpaired adults aged 45+ in two rural Midwestern U.S. communities, integrating environmental, social, behavioral, geospatial, and biological measures to evaluate exposure-related neurobiological and cognitive vulnerability. RESULTS: This approach demonstrated high feasibility and acceptability, achieving strong recruitment, retention, data completeness, and multidomain biomarker collection in rural community-based settings DISCUSSION: Pilot findings support the feasibility of neural exposome-informed research in rural U.S. communities and highlight its potential to advance prevention-oriented research on brain health and AD/ADRD.

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.

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.

The surge in medical imaging has spurred the development of vision-language models (VLMs) to alleviate radiologist workloads. However, clinical deployment is hindered by the lack of meaningful evaluation frameworks. Current metrics - ranging from semantic similarity to large language model (LLM) based judges - often fail to distinguish between clinically trivial and critical discrepancies, poorly reflecting real-world clinical judgment. To address this, we introduce DISCERN (Discordance and Significance-aware Entity-level Radiology Report Comparison). DISCERN is a significance-aware framework that weighs report errors based on their potential impact on patient care. Our results demonstrate that DISCERN powered by closed source LLMs aligns more closely with expert radiologist assessments than traditional metrics or current LLM evaluators, providing a more interpretable and clinically relevant benchmark. By modeling radiologist prioritization and entity-level feedback, DISCERN facilitates targeted model refinement and ensures the safer integration of generative AI into clinical workflows.

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.

Hypoxia-targeted BOLD MRI is a novel technique, which probes oxygenation physiology in response to a controlled transient hypoxia stimulus. In glioblastoma, the signal response is spatially and temporally heterogeneous. We developed a voxel-wise temporal decomposition framework for hypoxia-targeted BOLD MRI that separates the arrival of responses, transition phases, and steady state during controlled isocapnic hypoxia. Twenty healthy controls underwent 3-T BOLD MRI during a double hypoxic step challenge to establish a normative reference. Three patients with newly diagnosed glioblastoma were included as proof-of-concept cases. For each voxel, we estimated response arrival delay (Delaycorr), delay to plateau, delay to return and an O2-normalized steady-state response (HypoxiaSS). Healthy-control maps were used to construct a voxel-wise normative atlas and, for HypoxiaSS, a global-response-adjusted model for patient deviation mapping. In healthy controls, HypoxiaSS showed lower supratentorial between-subject variabilitythan both whole-stimulus comparators (coefficient of variation: 1.77 versus 2.36 for Hypoxiaavg) and higher voxel-level step-to-step agreement (ICC(2,1): median 0.951 versus 0.792 for Hypoxiaavg). Whole-stimulus averaging exhibited a systematic step-2 signal amplification present in 19 of 20 subjects, which was absent from HypoxiaSS. Asingle global response scalar explained a median 72.5% of voxel-wise between-subject variance in HypoxiaSS. In proof-of-concept patient analyses, G-adjusted HypoxiaSS deviation maps and timing maps identified spatially coherentabnormalities that were partly complementary and extended beyond conventional MRI-defined lesion margins.Temporal decomposition improves the stability and interpretability of hypoxia-targeted BOLD MRI and provides a practical framework for population-referenced physiological mapping and atlas-based deviation mapping in glioblastoma.

Andes orthohantavirus (ANDV), the primary etiological agent of hantavirus pulmonary syndrome (HPS) in South America, is uniquely capable of limited human-to-human transmission, posing a significant challenge for outbreak control. Recent events, including the 2018-2019 Epuyen outbreak and the 2026 MV Hondius incident, underscore the need for rapid, lineage-specific molecular diagnostics. In this study, we present an artificial intelligence (AI)-driven framework for the design of diagnostic primers targeting the S genomic segment of the Epuyen lineage. Using an evolutionary algorithm integrated with thermodynamic evaluation via Primer3Plus, candidate primers were optimized to maximize classification accuracy while satisfying stringent biochemical constraints. The resulting primer set enables amplification of lineage-specific regions suitable for molecular characterization and surveillance. In silico validation demonstrates that the proposed primers achieve perfect discrimination between 2026 outbreak sequences and other ANDV variants. Furthermore, in silico comparison with standard protocol-based primers reveals substantially reduced sensitivity and specificity in the latter, highlighting the limitations of static diagnostic designs when applied to evolving viral populations. Overall, this work demonstrates that AI-assisted primer design provides a robust and adaptable strategy to improve viral detection, enhance outbreak tracking, and support timely public health interventions. Integrating computational optimization into diagnostic development is essential for strengthening preparedness against emerging zoonotic threats.

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.

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: 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.