Geraint Rees

Background: Ischaemic stroke, a leading cause of death and disability, relies on neuroimaging for characterising the anatomical pattern of injury. Diffusion-weighted MRI (DWI) provides the most anatomically specific signal in acute ischaemic stroke but poses substantial challenges for automated lesion segmentation due to susceptibility artefacts, lesion morphological heterogeneity, comorbid pathology, instrumental variability, and limited labelled data. Current U-Net-based models therefore underperform, a problem accentuated by evaluation metrics that neglect anatomical, subpopulation and acquisition-dependent variability. Methods: We train 3D vision transformer–based segmentation models on a multi-site DWI dataset comprising 3563 annotated lesion-positive and 6900 lesion-negative volumes, using balanced cross-validation splits. We compare these models with U-Net baselines and an nnU-Net configuration under harmonised augmentation and training schemes, and introduce an evaluation framework that quantifies fidelity, anatomical precision, robustness to image corruption and equity across demographic and lesion-defined subtypes. Results: Here, we show that transformer-based models with our proposed control-image regularisation achieve higher segmentation performance than U-Net-based approaches on clinically realistic data, while substantially reducing false positives in lesion-negative images. They exhibit more stable performance across lesion sizes, anatomical territories, image quality and patient subgroups, indicating improved epistemic equity relative to convolutional architectures. Conclusions: This work reconciles model expressivity with domain-specific challenges and redefines performance benchmarks to prioritise equity and generalisability–critical for personalised medicine and mechanistic research. These findings establish vision transformer architectures, combined with equity-aware validation, as a powerful approach for ischaemic stroke lesion segmentation on DWI.

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.

Hyperconnectivity in functional brain networks occurs decades before disease onset in Huntington’s disease. However, the biological mechanisms remain unknown. We investigate connectivity in Huntington’s disease using Morphometric INverse Divergence (MIND) in three Huntington’s disease cohorts (N = 512) spanning from two decades before the onset of symptoms through to functional decline. Here, we identify stage-specific profiles, with hyperconnectivity 22 years from predicted motor onset, progressing to hypoconnectivity through the late premanifest and manifest stages, showing that hypoconnectivity is correlated with neurofilament light concentrations. To understand the biological mechanisms, we investigate associations with cortical organization principles including disease epicentres and cell-autonomous systems, in particular neurotransmitter distribution. The contribution from disease epicentres is limited to late premanifest while cell-autonomous associations are demonstrated across the Huntington’s disease lifespan. Specific relationships to cholinergic and serotoninergic systems localized to granular and infragranular cortical layers are identified, consistent with serotoninergic layer 5a neuronal vulnerability previously identified in post-mortem brains.

This viewpoint paper addresses the ongoing challenges and opportunities within the data-for-health ecosystem, drawing insights from a multistakeholder workshop. Despite notable progress in the digitization of health care systems, data sharing and interoperability remain limited, so the full potential of health care data is not realized. There is a critical need for data ecosystems that can enable the timely, safe, efficient, and sustainable collection and sharing of health care data. However, efforts to meet this need face risks related to privacy, data protection, security, democratic governance, and exclusion. Key challenges include poor interoperability, inconsistent approaches to data governance, and concerns about the commodification of data. While emerging platforms such as social media play a growing role in gathering and sharing health information, their integration into formal data systems remains limited. A robust and secure data-for-health ecosystem requires stronger frameworks for data governance, interoperability, and citizen engagement to build public trust. This paper argues that reframing health care data as a common good, improving the transparency of data acquisition and processing, and promoting the use of application programming interfaces (APIs) for real-time data access are essential to overcoming these challenges. In addition, it highlights the need for international norms and standards guided by multisector leadership, given the multinational nature of data sharing. Ultimately, this paper emphasizes the need to balance risks and opportunities to create a socially acceptable, secure, and effective data-sharing ecosystem in health care.

Huntington’s disease (HD) is an autosomal dominant neurodegenerative disease with the age at which characteristic symptoms manifest strongly influenced by inherited HTT CAG length. Somatic CAG expansion occurs throughout life and understanding the impact of somatic expansion on neurodegeneration is key to developing therapeutic targets. In 57 HD gene expanded (HDGE) individuals, ~23 years before their predicted clinical motor diagnosis, no significant decline in clinical, cognitive or neuropsychiatric function was observed over 4.5 years compared with 46 controls (false discovery rate (FDR) > 0.3). However, cerebrospinal fluid (CSF) markers showed very early signs of neurodegeneration in HDGE with elevated neurofilament light (NfL) protein, an indicator of neuroaxonal damage (FDR = 3.2 × 10⁻¹²), and reduced proenkephalin (PENK), a surrogate marker for the state of striatal medium spiny neurons (FDR = 2.6 × 10⁻³), accompanied by brain atrophy, predominantly in the caudate (FDR = 5.5 × 10⁻¹⁰) and putamen (FDR = 1.2 × 10⁻⁹). Longitudinal increase in somatic CAG repeat expansion ratio (SER) in blood was a significant predictor of subsequent caudate (FDR = 0.072) and putamen (FDR = 0.148) atrophy. Atypical loss of interruption HTT repeat structures, known to predict earlier age at clinical motor diagnosis, was associated with substantially faster caudate and putamen atrophy. We provide evidence in living humans that the influence of CAG length on HD neuropathology is mediated by somatic CAG repeat expansion. These critical mechanistic insights into the earliest neurodegenerative changes will inform the design of preventative clinical trials aimed at modulating somatic expansion. ClinicalTrials.gov registration: NCT06391619.

Background: The presence of bias in artificial intelligence has garnered increased attention, with inequities in algorithmic performance being exposed across the fields of criminal justice, education, and welfare services. In health care, the inequitable performance of algorithms across demographic groups may widen health inequalities. Objective: Here, we identify and characterize bias in cardiology algorithms, looking specifically at algorithms used in the management of heart failure. Methods: Stage 1 involved a literature search of PubMed and Web of Science for key terms relating to cardiac machine learning (ML) algorithms. Papers that built ML models to predict cardiac disease were evaluated for their focus on demographic bias in model performance, and open-source data sets were retained for our investigation. Two open-source data sets were identified: (1) the University of California Irvine Heart Failure data set and (2) the University of California Irvine Coronary Artery Disease data set. We reproduced existing algorithms that have been reported for these data sets, tested them for sex biases in algorithm performance, and assessed a range of remediation techniques for their efficacy in reducing inequities. Particular attention was paid to the false negative rate (FNR), due to the clinical significance of underdiagnosis and missed opportunities for treatment. Results: In stage 1, our literature search returned 127 papers, with 60 meeting the criteria for a full review and only 3 papers highlighting sex differences in algorithm performance. In the papers that reported sex, there was a consistent underrepresentation of female patients in the data sets. No papers investigated racial or ethnic differences. In stage 2, we reproduced algorithms reported in the literature, achieving mean accuracies of 84.24% (SD 3.51%) for data set 1 and 85.72% (SD 1.75%) for data set 2 (random forest models). For data set 1, the FNR was significantly higher for female patients in 13 out of 16 experiments, meeting the threshold of statistical significance (–17.81% to –3.37%; P<.05). A smaller disparity in the false positive rate was significant for male patients in 13 out of 16 experiments (–0.48% to +9.77%; P<.05). We observed an overprediction of disease for male patients (higher false positive rate) and an underprediction of disease for female patients (higher FNR). Sex differences in feature importance suggest that feature selection needs to be demographically tailored. Conclusions: Our research exposes a significant gap in cardiac ML research, highlighting that the underperformance of algorithms for female patients has been overlooked in the published literature. Our study quantifies sex disparities in algorithmic performance and explores several sources of bias. We found an underrepresentation of female patients in the data sets used to train algorithms, identified sex biases in model error rates, and demonstrated that a series of remediation techniques were unable to address the inequities present.

UNSTRUCTURED This viewpoint paper addresses the ongoing challenges and opportunities within the data-for-health ecosystem, drawing insights from a multistakeholder workshop. Despite notable progress in the digitization of health care systems, data sharing and interoperability remain limited, so the full potential of health care data is not realized. There is a critical need for data ecosystems that can enable the timely, safe, efficient, and sustainable collection and sharing of health care data. However, efforts to meet this need face risks related to privacy, data protection, security, democratic governance, and exclusion. Key challenges include poor interoperability, inconsistent approaches to data governance, and concerns about the commodification of data. While emerging platforms such as social media play a growing role in gathering and sharing health information, their integration into formal data systems remains limited. A robust and secure data-for-health ecosystem requires stronger frameworks for data governance, interoperability, and citizen engagement to build public trust. This paper argues that reframing health care data as a common good, improving the transparency of data acquisition and processing, and promoting the use of application programming interfaces (APIs) for real-time data access are essential to overcoming these challenges. In addition, it highlights the need for international norms and standards guided by multisector leadership, given the multinational nature of data sharing. Ultimately, this paper emphasizes the need to balance risks and opportunities to create a socially acceptable, secure, and effective data-sharing ecosystem in health care.