Contemporary clinical management relies on a diagnostic label as the primary guide to treatment. However, individual patients’ lived experiences vary more widely than standard diagnostic categories reflect. This is especially true for functional bowel disorders (FBDs), a heterogeneous and challenging group of syndromes where no definitive diagnostic tests, clinical biomarkers, or universally effective treatments exist. Characterising the link between disease and lived experience - in the face of marked patient heterogeneity - requires deep phenotyping of the interactions between multiple characteristics, plausibly achievable only with complex modelling approaches. In a large patient cohort (n = 1175), we developed a machine learning and Bayesian generative graph framework to better understand the lived experience of FBDs. Iterating through 59 factors available from routine clinical care, spanning patient demography, diagnosis, symptomatology, life impact, mental health indices, healthcare access requirements, COVID-19 impact, and treatment effectiveness, machine models were used to quantify the predictive fidelity of one feature from the remainder. Bayesian stochastic block models were used to delineate the network community structure underpinning the heterogeneous lived experience of FBDs. Machine models quantified patient personal health rating (R2 0.35), anxiety and depression severity (R2 0.54), employment status (balanced accuracy 96%), frequency of healthcare attendance (R2 0.71), and patient-reported treatment effectiveness variably (R2 range 0.08–0.41). Contrary to the view of many healthcare professionals, the greatest model predictors of patient-reported health and quality of life were life impact, mental well-being, employment status, and age, rather than diagnostic group or symptom severity. Patients responsive to one treatment were more likely to respond to another, leaving many others refractory to all. Clinical assessment of patients with FBDs should be less concerned with diagnostic classification than with the wider life impact of illness, including mental health and employment. The stratification of treatment response (and resistance) has implications for clinical practice and trial design, necessitating further research.

The gold standard in the treatment of ischaemic stroke is set by evidence from randomized controlled trials, typically using simple estimands of presumptively homogeneous populations. Yet the manifest complexity of the brain's functional, connective, and vascular architectures introduces heterogeneities that violate the underlying statistical premisses, potentially leading to substantial errors at both individual and population levels. The counterfactual nature of interventional inference renders quantifying the impact of this defect difficult. Here we conduct a comprehensive series of semi-synthetic, biologically plausible, virtual interventional trials across 100M+ distinct simulations. We generate empirically grounded virtual trial data from large-scale meta-analytic connective, functional, genetic expression, and receptor distribution data, with high-resolution maps of 4K+ acute ischaemic lesions. Within each trial, we estimate treatment effects using models varying in complexity, in the presence of increasingly confounded outcomes and noisy treatment responses. Individualized prescriptions inferred from simple models, fitted to unconfounded data, are less accurate than those from complex models, even when fitted to confounded data. Our results indicate that complex modelling with richly represented lesion data may substantively enhance individualized prescriptive inference in ischaemic stroke.

Accurate outcome prediction in acute ischaemic stroke is critical to clinical decision-making and optimal trial design. Predictive performance may, however, vary inequitably across multidimensionally defined patient subpopulations shaped by the intersection of diverse demographic, clinical, and anatomical factors such as the lesion distribution. Here we investigate the variation in achievable predictive performance— the epistemic equity of the model—across a multi-site cohort of 4083 richly characterised stroke episodes. Flexible machine learning models of demographic, clinical, and diffusion-weighted imaging data were used to predict an array of care needs and clinical outcomes. To identify systematic multidimensional subpopulations, deep representation learning was used to derive low-dimensional latent projections of high-dimensional clinical and imaging data. The distinct, consistent, interpretable multidimensional subpopulations thereby identified were used to quantify model epistemic equity across the population. Predictive performance varied substantially across subpopulations, revealing disparities invisible to simple, unidimensional measures of equity. Combining deep representation learning with stratified performance evaluation, we demonstrate the need for expressive, multidimensional representations in assuring the equity of predictive models in acute stroke care and research.

INTRODUCTION:Cannabinoids are being explored as potential treatments for gastroparesis due to their anti-emetic, gastric motility modulation, appetite stimulation, and analgesic properties coupled with their increasing use due to legalization in many states. Although these theoretical benefits are promising, clinical evidence remains limited. This study aimed to evaluate the effects of cannabis use on clinical outcomes and healthcare utilization in patients with gastroparesis using large-scale real-world data. METHODS:We conducted a cohort study using the TriNetX research network to identify US adults (≥18 years) with gastroparesis. From an initial cohort of 119 million individuals patients were stratified into cannabis users and non-users (controls). Propensity score matching (1:1) accounted for demographics, body mass index, comorbidities, laboratory parameters, and treatments. Primary outcomes included emergency department visits, hospitalizations, and esophagogastroduodenoscopy rates. RESULTS:Among 41,374 patients with gastroparesis, cannabis users (n = 20,687) and non-users (n = 20,687) were propensity-matched. Cannabis users were younger with higher rates of diabetes, mood/anxiety disorders, elevated hemoglobin A1c, and opioid use ( P < 0.001). Cannabis use was associated with increased emergency department visits (adjusted odds ratio [aOR] = 1.73, 95% confidence interval [CI]: 1.66-1.80) and hospitalizations (aOR = 1.44, 95% CI: 1.39-1.50) but reduced esophagogastroduodenoscopy utilization (aOR = 0.93, 95% CI: 0.88-0.98). DISCUSSION:Cannabis use in patients with gastroparesis seems to increase healthcare utilization. These findings underscore the need to carefully assess the risks and benefits of cannabis in gastroparesis management. Prospective studies are essential to evaluate cannabinoids' efficacy and safety in this context.

Complex heterogeneity is a major obstacle to clinical and scientific progress in neuro-oncology, resolvable only with highly expressive models powered by large-scale data and high-performance compute. We sought to de- velop a comprehensive multi-modal modelling framework with sufficient expressivity to capture heterogeneity distributed across the multiple characteristics that distinguish individual glioma. Evaluating a multi-site cohort of 4908 glioma patients, we employed multi-modal spatial autoencoders to derive succinct, spatially organised latent representations of anatomically organised patterns of heterogeneity. We quantified the prognostic and diagnostic utility of these representations within XGBoost discriminative models and compared their performance against conventional brain tumour featurisation radiomics and VASARI. The best prognostic models combined our deep autoencoder representations, VASARI scores, and radiomics, with histology (test set c-index=0.74). Prognostic models with the spatial autoencoder representations alone achieved a c-index of 0.71, surpassing both VASARI (c-index=0.69) and radiomics (c-index=0.69). The best diag- nostic model combined our deep autoencoder representations, VASARI scores, and radiomics, successfully dis- tinguishing glioblastoma from oligodendroglioma and astrocytoma with a test set balanced accuracy of 92.6%. The deep autoencoder alone could predict diagnosis with a balanced accuracy of 82.1%, while VASARI alone achieved 81.7% and radiomics alone reached 80.5%. Deep representations drawn from multimodal imaging data successfully capture heterogeneity opaque to con- ventional glioma characterisation and carry potentially useful predictive and diagnostic signals. These models may facilitate better patient care personalisation with more robust image-guided diagnostics and long-term outcome monitoring. Future research will harness these models to uncover new mechanistic insights.

Brain tumour MRI typically requires both pre- and post-contrast imaging, but gadolinium is not always desirable (frequent follow-up, renal impairment, allergy, paediatric patients). We developed and validated a deep learning model to predict tumour contrast enhancement from non-contrast MRI alone. We assembled 11,089 brain MRI studies (2006-2024) from 10 datasets across four countries and three continents, spanning adult and paediatric populations with glioma, meningioma, metastases, and post-resection appearances. Three architectures were trained to detect and segment enhancing tumour from T1w, T2w and FLAIR alone. Performance was assessed in a 1,109-study held-out test set (primary endpoint: patient-level enhancement detection; secondary: voxel-level Dice). Eleven expert radiologists attempted the same task on a 564-case subset (100 cases each), blinded to history, prior imaging, and referral. The best model, nnU-Net, achieved 83.0

Two of the most well-studied types of reasoning are analogical reasoning (AR) and deductive reasoning (DR). Yet, our understanding of the relationship between reasoning abilities and their neuroanatomical basis remains surprisingly limited. We aimed to conduct fine-grained anatomical mapping of performance on tests of AR, DR and fluid intelligence (Gf), in a large sample of patients with unilateral focal frontal or posterior lesions and healthy controls. We assessed 247 prospectively recruited patients using two new tests: the Analogical Reasoning Test (ART) and the Deductive Reasoning Test (DRT); and the best-established measure of Gf: Raven's Advanced Progressive Matrices (RAPM). Non-parametric Bayesian stochastic block modelling was used to reveal the community structure of lesion deficit networks, disentangling functional from confounding pathological distributed effects. ART and DRT performance was significantly impaired in patients with frontal lesions [ART: F(2,238) = 18.93; P < 0.001; Frontal group worse than Posterior group and healthy controls, both P < 0.001; DRT: F(2,387) = 18.491; P < 0.001; Frontal group worse than healthy controls, P < 0.01]. Right frontal effects were evident on both tests. Thus, on the ART, right frontal patients were more impaired than left (P < 0.05). On the DRT, right frontal patients were more impaired than left frontal patients on questions with indeterminate solutions (P < 0.05) but not on questions with determinate ones. Non-parametric Bayesian stochastic block modelling implicated a right frontal network in ART and DRT performance. Strikingly, we found that this network was also implicated in performance on RAPM. Our study represents the most robust investigation of AR and DR in the focally injured brain. Our findings imply that a right frontal network is critical. The ART and DRT appear to be promising new clinical tests, capable of evaluating reasoning abilities and identifying right frontal lobe dysfunction.

Ischaemic stroke, a leading cause of death and disability, critically relies on neuroimaging for characterising the anatomical pattern of injury. Diffusion-weighted imaging (DWI) provides the highest expressivity in ischemic stroke but poses substantial challenges for automated lesion segmentation: susceptibility artefacts, morphological heterogeneity, age-related comorbidities, time-dependent signal dynamics, instrumental variability, and limited labelled data. Current U-Net-based models therefore underperform, a problem accentuated by inadequate evaluation metrics that focus on mean performance, neglecting anatomical, subpopulation, and acquisition-dependent variability. Here, we present a high-performance DWI lesion segmentation tool addressing these challenges through optimized vision transformer-based architectures, integration of 3563 annotated lesions from multi-site data, and algorithmic enhancements, achieving state-of-the-art results. We further propose a novel evaluative framework assessing model fidelity, equity (across demographics and lesion subtypes), anatomical precision, and robustness to instrumental variability, promoting clinical and research utility. This work advances stroke imaging by reconciling model expressivity with domain-specific challenges and redefining performance benchmarks to prioritize equity and generalizability, critical for personalized medicine and mechanistic research.

The quantification of cognitive powers rests on identifying a behavioural task that depends on them. Such dependence cannot be assured, for the powers a task invokes cannot be experimentally controlled or constrained a priori, resulting in unknown vulnerability to failure of specificity and generalisability. Evaluating a compact version of Raven's Advanced Progressive Matrices (RAPM), a widely used clinical test of fluid intelligence, we show that LaMa, a self-supervised artificial neural network trained solely on the completion of partially masked images of natural environmental scenes, achieves representative human-level test scores a prima vista, without any task-specific inductive bias or training. Compared with cohorts of healthy and focally lesioned participants, LaMa exhibits human-like variation with item difficulty, and produces errors characteristic of right frontal lobe damage under degradation of its ability to integrate global spatial patterns. LaMa's narrow training and limited capacity suggest matrix-style tests may be open to computationally simple solutions that need not necessarily invoke the substrates of reasoning.

Objective Functional bowel disorders (FBDs) are multi-dimensional diseases varying in demographics, symptomology, lifestyle, mental health, and susceptibility to treatment. The patient lived experience is an integration of these factors, best understood with appropriately multivariate models. Methods In a large patient cohort (n=1175), we developed a machine learning framework to better understand the lived experience of FBDs. Iterating through 59 factors available from routine clinical care, spanning patient demography, diagnosis, symptomatology, life-impact, mental health indices, healthcare access requirements, COVID-19 impact, and treatment effectiveness, machine models were used to quantify the predictive fidelity of one feature from the remainder. Bayesian stochastic block models were used to delineate the network community structure underpinning the lived experience of FBDs. Results Machine models quantified patient personal health rating (R2 0.35), anxiety and depression severity (R2 0.54), employment status (balanced accuracy 96%), frequency of healthcare attendance (R2 0.71), and patient-reported treatment effectiveness variably (R2 range 0.08-0.41). Contrary to the view of many healthcare professionals, the greatest determinants of patient-reported health and quality-of-life were life-impact, mental wellbeing, employment status, and age, rather than diagnostic group and symptom severity. Patients responsive to one treatment were more likely to respond to another, leaving many others refractory to all. Conclusions The assessment of patients with FBDs should be less concerned with diagnostic classification than with the wider life impact of illness, including mental health and employment. The stratification of treatment response (and resistance) has implications for clinical practice and trial design, in need of further research. What is known? What is new here? ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement JKR was supported by the Medical Research Council (MR/X00046X/1). PN is supported by the Wellcome Trust (213038/Z/18/Z) and the UCLH NIHR Biomedical Research Centre. The PERSPECTIVE study was funded by MacGregor Healthcare Ltd. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study was approved by local institutional review board and conducted in accordance with the Declaration of Helsinki. The Health Research Authority approved this study prior to commencement. REC reference 21/SW/0086 (IRAS ID 296856). I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes All code will be made publicly available upon publication at https://github.com/jamesruffle/perspective-ai. Trained model weights are available upon request. Data and code availability is in line with UK government policy on open-source code. Patient data are not available for dissemination under the ethical framework that governs its use.

BACKGROUND:A high prevalence of disorders of gut-brain interaction (DGBI) exist in patients with hypermobile Ehlers-Danlos Syndrome (hEDS) and hypermobility spectrum disorders (HSD). However, it is unknown if clusters of hEDS/HSD patients exist which overlap with different DGBIs and whether this overlap influences presence of comorbidities and quality of life. We aimed to study these knowledge gaps. METHODS:A prospectively collected hEDS/HSD cohort of 1044 individuals were studied. We undertook Uniform Manifold Approximation and Projection-enabled (UMAP) dimension reduction to create a representation of nonlinear interactions between hEDS/HSD and DGBIs, from which individuals were stratified into clusters. Somatization, Postural Tachycardia Syndrome (PoTS), autonomic symptoms, psychological factors and quality of life were statistically compared between clusters. KEY RESULTS:The mean age of patients was 40 ± 13.2 years; 87.8% were female. Patients segregated into three clusters: Cluster 0 (n = 466): hEDS/HSD+ functional foregut disorders (FFD) + irritable bowel syndrome (IBS); Cluster 1 (n = 180): hEDS/HSD+ IBS and Cluster 2 (n = 337): hEDS/HSD alone. In cluster 0, we demonstrated increased somatization (p <0.0001), anxiety (p <0.0001), depression (p <0.0001), PoTS prevalence (p = 0.003), autonomic symptoms (p <0.0001) and reduced quality of life (p <0.0001) compared to cluster 2. Cluster 0 had greater comorbidity burden than cluster 1. CONCLUSIONS:Within hEDS/HSD, subgroups exist with a high prevalence of FFD and IBS. These subgroups have a higher prevalence of psychological disorders, dysautonomia and poorer quality of life compared with hEDS/HSD alone. Further research should focus on healthcare utilization, management and prognosis in hEDS/HSD and DGBI overlap.

The architecture of the brain is too complex to be intuitively surveyable without the use of compressed representations that project its variation into a compact, navigable space. The task is especially challenging with high-dimensional data, such as gene expression, where the joint complexity of anatomical and transcriptional patterns demands maximum compression. The established practice is to use standard principal component analysis (PCA), whose computational felicity is offset by limited expressivity, especially at great compression ratios. Employing whole-brain, voxel-wise Allen Brain Atlas transcription data, here we systematically compare compressed representations based on the most widely supported linear and non-linear methods-PCA, kernel PCA, non-negative matrix factorisation (NMF), t-stochastic neighbour embedding (t-SNE), uniform manifold approximation and projection (UMAP), and deep auto-encoding-quantifying reconstruction fidelity, anatomical coherence, and predictive utility across signalling, microstructural, and metabolic targets, drawn from large-scale open-source MRI and PET data. We show that deep auto-encoders yield superior representations across all metrics of performance and target domains, supporting their use as the reference standard for representing transcription patterns in the human brain.

Motivation: Early detection of low-grade glioma (LGG) malignant transformation (MT) is vital for treatment decisions, prognosis, quality of life and patient-centered care. Goal(s): To develop non-linear machine learning models using XGBoost algorithm to predict overall survival using clinical, molecular, genetic and radiomic data at MT. Approach: 553 LGGs with histology and MRI underwent in-house tumour segmentation pipeline with radiomic feature extraction and masked disconnectome of map components. Results: XGB Classifier model predicted OS > 5 years from MT with an accuracy of 64%. Age, IDH1 mutation, 1p/19q co-deletion, regularity of tumour shape, and disconnectome-related perilesional components were most predictive of survival. Impact: Understanding malignant transformation of low-grade gliomas is crucial for research and the development of new treatment strategies. Defining the radiological features at malignant transformation allows for a timely shift in the treatment plan with potential to improve repsonse to therapy.

Abstract AIMS Residual tumour burden after surgery in GBM patients is a prognostic imaging biomarker, often involving the area of T2 signal elevation surrounding the tumour core, histopathologically representing infiltrative oedema with tumour cells. We present the first stage of a model to improve the prediction GBM recurrence by specifically investigating the peri-lesional oedema compartment. METHOD UK-wide sample of histologically proven IDHwt GBMs with MRIs at the time of diagnosis underwent state-of- the-art in-house tumour segmentation pipeline and radiomic feature extraction of whole-tumour, enhancing, non-enhancing and oedema components, and masked disconnectome map components. Of these, early GBM recurrence, within 6 months after completion of radiotherapy, was curated and distribution mapped across the brain. RESULTS 1464 patients were identified: male = 901, median age 63 years, average overall survival 597 days. The radiomic features: Compactness 1 and 2, Surface Area, Spherical Disproportion, Sphericity, and Maximum 2D Diameter column of the disconnectome-related perilesional oedema compartment were most predictive of survival outcome (p value <0.01). Of these, 47 patients with average overall survival 357 days and average time from end of radiotherapy to progression of 131 days were identified. CONCLUSION Our early results reveal the relationship between perilesional oedema MRI features and survival, with disconnectome-related perilesional oedema features most predictive. This suggests that extended brain network aberrations caused by gliomas may also influence tumour recurrence. Further work will focus on the development of more robust predictive models of tumour recurrence that have potential for significant impact on surgical and radiotherapy planning.

Intracranial pressure (ICP) is a physiological parameter that conventionally requires invasive monitoring for accurate measurement. Utilising multivariate predictive models, we sought to evaluate the utility of non-invasive, widely accessible MRI biomarkers in predicting ICP and their reversibility following cerebrospinal fluid (CSF) diversion. The retrospective study included 325 adult patients with suspected CSF dynamic disorders who underwent brain MRI scans within three months of elective 24-h ICP monitoring. Five MRI biomarkers were assessed: Yuh sella grade, optic nerve vertical tortuosity (VT), optic nerve sheath distension, posterior globe flattening and optic disc protrusion (ODP). The association between individual biomarkers and 24-h ICP was examined and reversibility of each following CSF diversion was assessed. Multivariate models incorporating these radiological biomarkers were utilised to predict 24-h median intracranial pressure. All five biomarkers were significantly associated with median 24-h ICP (p < 0.0001). Using a pair-wise approach, the presence of each abnormal biomarker was significantly associated with higher median 24-h ICP (p < 0.0001). On multivariate analysis, ICP was significantly and positively associated with Yuh sella grade (p < 0.0001), VT (p < 0.0001) and ODP (p = 0.003), after accounting for age and suspected diagnosis. The Bayesian multiple linear regression model predicted 24-h median ICP with a mean absolute error of 2.71 mmHg. Following CSF diversion, we found pituitary sella grade to show significant pairwise reversibility (p < 0.001). ICP was predicted with clinically useful precision utilising a compact Bayesian model, offering an easily interpretable tool using non-invasive MRI data. Brain MRI biomarkers are anticipated to play a more significant role in the screening, triaging, and referral of patients with suspected CSF dynamic disorders.

Abstract AIMS The VASARI (Visually AcceSAble Rembrandt Images) MRI feature set is a quantitative system designed to standardize glioma imaging descriptions. Though effective, deriving VASARI is time-consuming and therefore seldom used in clinical practice. This is however a problem that performant machine-learning software could plausibly automate. METHOD In 100 patients, two consultant neuroradiologists independently quantified VASARI features. In parallel, we developed VASARI-auto, an automated VASARI-labelling software applied to both open-source hand-segmented lesion masks and our openly available tumour segmentation model. We quantified: 1) agreement across neuroradiologists and VASARI-auto; 2) calibration of patient equity; 3) an economic workforce analysis; and 4) fidelity in predicting patient overall survival. Neuroradiologists were blinded to VASARI-auto software development and evaluations, and software developers were blinded to neuroradiologist labelling. RESULTS Segmentation and VASARI-auto were equally performant regardless of age or sex (mean segmentation Dice coefficient 0.947). A modest inter-rater variability between neuroradiologists (mean Cohen’s Kappa 0.49) was comparable between neuroradiologists and VASARI-auto (mean Cohen’s Kappa 0.41), with higher agreement across VASARI-auto methods (mean Cohen’s Kappa 0.94). The time taken for neuroradiologists to derive VASARI features was substantially higher than VASARI-auto (mean time per case 317 vs. 3 seconds, p<0.0001). A hospital workforce analysis of our centre forecast that three years of VASARI glioma featurisation would demand 771 consultant neuroradiologist workforce hours (£40,789) or 8.6 hours of computing time (£3.76 of power) with VASARI-auto. The best-performing survival model utilised VASARI-auto features (R2 0.25), opposed to those derived by consultant neuroradiologists (R2 0.21). CONCLUSION VASARI-auto is a highly efficient automated labelling system for VASARI featurisation with equitable performance regardless of patient age or sex, a favourable economic profile if used as a decision support tool, and with non-inferior fidelity in downstream patient survival prediction. Future work should integrate such tools to enhance patient care.

Influential theories of complex behaviour invoke the notion of cognitive control modulated by conflict between counterfactual actions. Medial frontal cortex, notably the anterior cingulate cortex, has been variously posited as critical to such conflict detection, resolution, or monitoring, largely based on correlative data from functional imaging. Examining performance on the most widely used "conflict" task-Stroop-in a large cohort of patients with focal brain injury (N = 176), we compare anatomical patterns of lesion-inferred neural substrate dependence to those derived from functional imaging, meta-analytically summarised. Our results show that whereas performance is sensitive to the integrity of left lateral frontal regions implicated by functional imaging, it does not depend on medial frontal cortex, despite sampling adequate to reveal robust medial effects in the context of phonemic fluency. We suggest that medial frontal cortex is not critically invoked by Stroop and proceed to review the conceptual grounds for rejecting the core notion of conflict-driven cognitive control.

The VASARI MRI feature set is a quantitative system designed to standardise glioma imaging descriptions. Though effective, deriving VASARI is time-consuming and seldom used clinically. We sought to resolve this problem with software automation and machine learning. Using glioma data from 1172 patients, we developed VASARI-auto, an automated labelling software applied to open-source lesion masks and an openly available tumour segmentation model. Consultant neuroradiologists independently quantified VASARI features in 100 held-out glioblastoma cases. We quantified 1) agreement across neuroradiologists and VASARI-auto, 2) software equity, 3) an economic workforce analysis, and 4) fidelity in predicting survival. Tumour segmentation was compatible with the current state of the art and equally performant regardless of age or sex. A modest inter-rater variability between in-house neuroradiologists was comparable to between neuroradiologists and VASARI-auto, with far higher agreement between VASARI-auto methods. The time for neuroradiologists to derive VASARI was substantially higher than VASARI-auto (mean time per case 317 vs. 3 s). A UK hospital workforce analysis forecast that three years of VASARI featurisation would demand 29,777 consultant neuroradiologist workforce hours and >£1.5 ($1.9) million, reducible to 332 hours of computing time (and £146 of power) with VASARI-auto. The best-performing survival model utilised VASARI-auto features instead of those derived by neuroradiologists. VASARI-auto is a highly efficient and equitable automated labelling system, a favourable economic profile if used as a decision support tool, and non-inferior survival prediction. Future work should iterate upon and integrate such tools to enhance patient care.

Our knowledge of the organisation of the human brain at the population-level is yet to translate into power to predict functional differences at the individual-level, limiting clinical applications, and casting doubt on the generalisability of inferred mechanisms. It remains unknown whether the difficulty arises from the absence of individuating biological patterns within the brain, or from limited power to access them with the models and compute at our disposal.Here we comprehensively investigate the resolvability of such patterns with data and compute at unprecedented scale. Across 23 810 unique participants from UK Biobank, we systematically evaluate the predictability of 25 individual biological characteristics, from all available combinations of structural and functional neuroimaging data. Over 4526 GPU*hours of computation, we train, optimize, and evaluate out-of-sample 700 individual predictive models, including fully-connected feed-forward neural networks of demographic, psychological, serological, chronic disease, and functional connectivity characteristics, and both uni- and multi-modal 3D convolutional neural network models of macro- and micro-structural brain imaging.We find a marked discrepancy between the high predictability of sex (balanced accuracy 99.7%), age (mean absolute error 2.048 years, R2 0.859), and weight (mean absolute error 2.609Kg, R2 0.625), for which we set new state-of-the-art performance, and the surprisingly low predictability of other characteristics. Neither structural nor functional imaging predicted an individual's psychology better than the coincidence of common chronic disease (p<0.05). Serology predicted chronic disease (p<0.05) and was best predicted by it (p<0.001), followed by structural neuroimaging (p<0.05).Our findings suggest either more informative imaging or more powerful models will be needed to decipher individual level characteristics from the human brain. We make our models and code openly available.

INTRODUCTION: Disorders of gut-brain interaction (DGBI) are common in patients with hypermobile Ehlers-Danlos syndrome/hypermobility spectrum disorder (hEDS/HSD). Food is a known trigger for DGBI symptoms, which often leads to dietary alterations and, increasingly, nutrition support. We aimed to explore dietary behaviors and influencing factors in patients with hEDS/HSD. METHODS: In a cross-sectional study, patients with hEDS/HSD were recruited from Ehlers-Danlos Support UK (nontertiary) and tertiary neurogastroenterology clinics to complete questionnaires characterizing the following: dietary behaviors, nutrition support, DGBI (Rome IV), gastrointestinal symptoms, anxiety, depression, avoidant restrictive food intake disorder (ARFID), mast cell activation syndrome, postural tachycardia syndrome (PoTS), and quality of life. We used stepwise logistic regression to ascertain which factors were associated with dietary behaviors and nutrition support. RESULTS: Of 680 participants (95% female, median age 39 years), 62.1% altered their diet in the last year and 62.3% regularly skipped meals. Altered diet was associated with the following: reflux symptoms ( P < 0.001), functional dyspepsia ( P = 0.008), reported mast cell activation syndrome ( P < 0.001), and a positive screen for ARFID, specifically fear of eating and low interest ( P < 0.001). Approximately 31.7% of those who altered their diet required nutrition support. The strongest predictor of requiring nutrition support was a positive screen for ARFID, specifically fear of eating (OR: 4.97, 95% CI: 2.09–11.8, P < 0.001). DISCUSSION: Altered diet is very common in the patients with hEDS/HSD we studied and influenced by functional dyspepsia, reflux symptoms, and ARFID. Those with ARFID have a 4-fold increased risk of requiring nutrition support, and therefore, it is paramount that psychological support is offered in parallel with dietary support in the management of DGBI in hEDS/HSD.