Miguel O. Bernabeu

Jamie Burke, Justin Engelmann, Charlene Hamid et al.

Purpose: To develop an open-source, fully-automatic deep learning algorithm, DeepGPET, for choroid region segmentation in optical coherence tomography (OCT) data. Methods: We used a dataset of 715 OCT B-scans (82 subjects, 115 eyes) from 3 clinical studies related to systemic disease. Ground truth segmentations were generated using a clinically validated, semi-automatic choroid segmentation method, Gaussian Process Edge Tracing (GPET). We finetuned a UNet with MobileNetV3 backbone pre-trained on ImageNet. Standard segmentation agreement metrics, as well as derived measures of choroidal thickness and area, were used to evaluate DeepGPET, alongside qualitative evaluation from a clinical ophthalmologist. Results: DeepGPET achieves excellent agreement with GPET on data from 3 clinical studies (AUC=0.9994, Dice=0.9664; Pearson correlation of 0.8908 for choroidal thickness and 0.9082 for choroidal area), while reducing the mean processing time per image on a standard laptop CPU from 34.49s ($\pm$15.09) using GPET to 1.25s ($\pm$0.10) using DeepGPET. Both methods performed similarly according to a clinical ophthalmologist, who qualitatively judged a subset of segmentations by GPET and DeepGPET, based on smoothness and accuracy of segmentations. Conclusions: DeepGPET, a fully-automatic, open-source algorithm for choroidal segmentation, will enable researchers to efficiently extract choroidal measurements, even for large datasets. As no manual interventions are required, DeepGPET is less subjective than semi-automatic methods and could be deployed in clinical practice without necessitating a trained operator.

Justin Engelmann, Amos Storkey, Miguel O. Bernabeu

Image quality remains a key problem for both traditional and deep learning (DL)-based approaches to retinal image analysis, but identifying poor quality images can be time consuming and subjective. Thus, automated methods for retinal image quality scoring (RIQS) are needed. The current state-of-the-art is MCFNet, composed of three Densenet121 backbones each operating in a different colour space. MCFNet, and the EyeQ dataset released by the same authors, was a huge step forward for RIQS. We present QuickQual, a simple approach to RIQS, consisting of a single off-the-shelf ImageNet-pretrained Densenet121 backbone plus a Support Vector Machine (SVM). QuickQual performs very well, setting a new state-of-the-art for EyeQ (Accuracy: 88.50% vs 88.00% for MCFNet; AUC: 0.9687 vs 0.9588). This suggests that RIQS can be solved with generic perceptual features learned on natural images, as opposed to requiring DL models trained on large amounts of fundus images. Additionally, we propose a Fixed Prior linearisation scheme, that converts EyeQ from a 3-way classification to a continuous logistic regression task. For this task, we present a second model, QuickQual MEga Minified Estimator (QuickQual-MEME), that consists of only 10 parameters on top of an off-the-shelf Densenet121 and can distinguish between gradable and ungradable images with an accuracy of 89.18% (AUC: 0.9537). Code and model are available on GitHub: https://github.com/justinengelmann/QuickQual . QuickQual is so lightweight, that the entire inference code (and even the parameters for QuickQual-MEME) is already contained in this paper.

Yinghao Zhu, Junyi Gao, Zixiang Wang et al.

Large Language Models (LLMs) are increasingly deployed in medicine. However, their utility in non-generative clinical prediction, often presumed inferior to specialized models, remains under-evaluated, leading to ongoing debate within the field and potential for misuse, misunderstanding, or over-reliance due to a lack of systematic benchmarking. Our ClinicRealm study addresses this by benchmarking 15 GPT-style LLMs, 5 BERT-style models, and 11 traditional methods on unstructured clinical notes and structured Electronic Health Records (EHR), while also assessing their reasoning, reliability, and fairness. Key findings reveal a significant shift: for clinical note predictions, leading LLMs (e.g., DeepSeek-V3.1-Think, GPT-5) in zero-shot settings now decisively outperform finetuned BERT models. On structured EHRs, while specialized models excel with ample data, advanced LLMs (e.g., GPT-5, DeepSeek-V3.1-Think) show potent zero-shot capabilities, often surpassing conventional models in data-scarce settings. Notably, leading open-source LLMs can match or exceed proprietary counterparts. These results provide compelling evidence that modern LLMs are competitive tools for non-generative clinical prediction, particularly with unstructured text and offering data-efficient structured data options, thus necessitating a re-evaluation of model selection strategies. This research should serve as an important insight for medical informaticists, AI developers, and clinical researchers, potentially prompting a reassessment of current assumptions and inspiring new approaches to LLM application in predictive healthcare.

Justin Engelmann, Ana Villaplana-Velasco, Amos Storkey et al.

A retinal trait, or phenotype, summarises a specific aspect of a retinal image in a single number. This can then be used for further analyses, e.g. with statistical methods. However, reducing an aspect of a complex image to a single, meaningful number is challenging. Thus, methods for calculating retinal traits tend to be complex, multi-step pipelines that can only be applied to high quality images. This means that researchers often have to discard substantial portions of the available data. We hypothesise that such pipelines can be approximated with a single, simpler step that can be made robust to common quality issues. We propose Deep Approximation of Retinal Traits (DART) where a deep neural network is used predict the output of an existing pipeline on high quality images from synthetically degraded versions of these images. We demonstrate DART on retinal Fractal Dimension (FD) calculated by VAMPIRE, using retinal images from UK Biobank that previous work identified as high quality. Our method shows very high agreement with FD VAMPIRE on unseen test images (Pearson r=0.9572). Even when those images are severely degraded, DART can still recover an FD estimate that shows good agreement with FD VAMPIRE obtained from the original images (Pearson r=0.8817). This suggests that our method could enable researchers to discard fewer images in the future. Our method can compute FD for over 1,000img/s using a single GPU. We consider these to be very encouraging initial results and hope to develop this approach into a useful tool for retinal analysis.

Justin Engelmann, Alice D. McTrusty, Ian J. C. MacCormick et al.

Ultra-widefield (UWF) imaging is a promising modality that captures a larger retinal field of view compared to traditional fundus photography. Previous studies showed that deep learning (DL) models are effective for detecting retinal disease in UWF images, but primarily considered individual diseases under less-than-realistic conditions (excluding images with other diseases, artefacts, comorbidities, or borderline cases; and balancing healthy and diseased images) and did not systematically investigate which regions of the UWF images are relevant for disease detection. We first improve on the state of the field by proposing a DL model that can recognise multiple retinal diseases under more realistic conditions. We then use global explainability methods to identify which regions of the UWF images the model generally attends to. Our model performs very well, separating between healthy and diseased retinas with an area under the curve (AUC) of 0.9206 on an internal test set, and an AUC of 0.9841 on a challenging, external test set. When diagnosing specific diseases, the model attends to regions where we would expect those diseases to occur. We further identify the posterior pole as the most important region in a purely data-driven fashion. Surprisingly, 10% of the image around the posterior pole is sufficient for achieving comparable performance to having the full images available.

Justin Engelmann, Jamie Burke, Charlene Hamid et al.

Purpose: To develop Choroidalyzer, an open-source, end-to-end pipeline for segmenting the choroid region, vessels, and fovea, and deriving choroidal thickness, area, and vascular index. Methods: We used 5,600 OCT B-scans (233 subjects, 6 systemic disease cohorts, 3 device types, 2 manufacturers). To generate region and vessel ground-truths, we used state-of-the-art automatic methods following manual correction of inaccurate segmentations, with foveal positions manually annotated. We trained a U-Net deep-learning model to detect the region, vessels, and fovea to calculate choroid thickness, area, and vascular index in a fovea-centred region of interest. We analysed segmentation agreement (AUC, Dice) and choroid metrics agreement (Pearson, Spearman, mean absolute error (MAE)) in internal and external test sets. We compared Choroidalyzer to two manual graders on a small subset of external test images and examined cases of high error. Results: Choroidalyzer took 0.299 seconds per image on a standard laptop and achieved excellent region (Dice: internal 0.9789, external 0.9749), very good vessel segmentation performance (Dice: internal 0.8817, external 0.8703) and excellent fovea location prediction (MAE: internal 3.9 pixels, external 3.4 pixels). For thickness, area, and vascular index, Pearson correlations were 0.9754, 0.9815, and 0.8285 (internal) / 0.9831, 0.9779, 0.7948 (external), respectively (all p<0.0001). Choroidalyzer's agreement with graders was comparable to the inter-grader agreement across all metrics. Conclusions: Choroidalyzer is an open-source, end-to-end pipeline that accurately segments the choroid and reliably extracts thickness, area, and vascular index. Especially choroidal vessel segmentation is a difficult and subjective task, and fully-automatic methods like Choroidalyzer could provide objectivity and standardisation.

Jesse Phitidis, Alison Q. Smithard, William N. Whiteley et al.

White matter hyperintensities (WMH) and ischaemic stroke lesions (ISL) are imaging features associated with cerebral small vessel disease (SVD) that are visible on brain magnetic resonance imaging (MRI) scans. The development and validation of deep learning models to segment and differentiate these features is difficult because they visually confound each other in the fluid-attenuated inversion recovery (FLAIR) sequence and often appear in the same subject. We investigated six strategies for training a combined WMH and ISL segmentation model using partially labelled data. We combined privately held fully and partially labelled datasets with publicly available partially labelled datasets to yield a total of 2052 MRI volumes, with 1341 and 1152 containing ground truth annotations for WMH and ISL respectively. We found that several methods were able to effectively leverage the partially labelled data to improve model performance, with the use of pseudolabels yielding the best result.

Jamie Burke, Justin Engelmann, Charlene Hamid et al.

The choroid is a key vascular layer of the eye, supplying oxygen to the retinal photoreceptors. Non-invasive enhanced depth imaging optical coherence tomography (EDI-OCT) has recently improved access and visualisation of the choroid, making it an exciting frontier for discovering novel vascular biomarkers in ophthalmology and wider systemic health. However, current methods to measure the choroid often require use of multiple, independent semi-automatic and deep learning-based algorithms which are not made open-source. Previously, Choroidalyzer -- an open-source, fully automatic deep learning method trained on 5,600 OCT B-scans from 385 eyes -- was developed to fully segment and quantify the choroid in EDI-OCT images, thus addressing these issues. Using the same dataset, we propose a Robust, Resolution-agnostic and Efficient Attention-based network for CHoroid segmentation (REACH). REACHNet leverages multi-resolution training with domain-specific data augmentation to promote generalisation, and uses a lightweight architecture with resolution-agnostic self-attention which is not only faster than Choroidalyzer's previous network (4 images/s vs. 2.75 images/s on a standard laptop CPU), but has greater performance for segmenting the choroid region, vessels and fovea (Dice coefficient for region 0.9769 vs. 0.9749, vessels 0.8612 vs. 0.8192 and fovea 0.8243 vs. 0.3783) due to its improved hyperparameter configuration and model training pipeline. REACHNet can be used with Choroidalyzer as a drop-in replacement for the original model and will be made available upon publication.

Justin Engelmann, Diana Moukaddem, Lucas Gago et al.

Purpose: To investigate whether Fractal Dimension (FD)-based oculomics could be used for individual risk prediction by evaluating repeatability and robustness. Methods: We used two datasets: Caledonia, healthy adults imaged multiple times in quick succession for research (26 subjects, 39 eyes, 377 colour fundus images), and GRAPE, glaucoma patients with baseline and follow-up visits (106 subjects, 196 eyes, 392 images). Mean follow-up time was 18.3 months in GRAPE, thus it provides a pessimistic lower-bound as vasculature could change. FD was computed with DART and AutoMorph. Image quality was assessed with QuickQual, but no images were initially excluded. Pearson, Spearman, and Intraclass Correlation (ICC) were used for population-level repeatability. For individual-level repeatability, we introduce measurement noise parameter λ which is within-eye Standard Deviation (SD) of FD measurements in units of between-eyes SD. Results: In Caledonia, ICC was 0.8153 for DART and 0.5779 for AutoMorph, Pearson/Spearman correlation (first and last image) 0.7857/0.7824 for DART, and 0.3933/0.6253 for AutoMorph. In GRAPE, Pearson/Spearman correlation (first and next visit) was 0.7479/0.7474 for DART, and 0.7109/0.7208 for AutoMorph (all p<0.0001). Median λ in Caledonia without exclusions was 3.55\% for DART and 12.65\% for AutoMorph, and improved to up to 1.67\% and 6.64\% with quality-based exclusions, respectively. Quality exclusions primarily mitigated large outliers. Worst quality in an eye correlated strongly with λ (Pearson 0.5350-0.7550, depending on dataset and method, all p<0.0001). Conclusions: Repeatability was sufficient for individual-level predictions in heterogeneous populations. DART performed better on all metrics and might be able to detect small, longitudinal changes, highlighting the potential of robust methods.

Justin Engelmann, Miguel O. Bernabeu

Artificial Intelligence in medicine is traditionally limited by the lack of massive training datasets. Foundation models, pre-trained models that can be adapted to downstream tasks with small datasets, could alleviate this problem. Researchers at Moorfields Eye Hospital (MEH) proposed RETFound-MEH, a retinal foundation model trained on 900,000 images, including private hospital data. Recently, data-efficient DERETFound was proposed providing comparable performance while being trained on only 150,000 publicly available images. However, both these models required very substantial resources to train initially and are resource-intensive in downstream use. We propose a novel Token Reconstruction objective that we use to train RETFound-Green, a retinal foundation model trained using only 75,000 publicly available images and 400 times less compute. We estimate the cost of training RETFound-MEH and DERETFound at \$10,000 and \$14,000, respectively. RETFound-Green could be trained for less than \$100, with equally reduced environmental impact. RETFound-Green is also far more efficient in downstream use: it can be downloaded 14 times faster, computes vector embeddings 2.7 times faster which then require 2.6 times less storage space. Despite this, RETFound-Green does not perform systematically worse. In fact, on various task on three downstream datasets from Brazil, India and China, it performs best on 68 tasks out of 119 comparisons, versus 21 for DERETFound and 13 for RETFound-MEH. Our results suggest that RETFound-Green is a very efficient, high-performance retinal foundation model. We anticipate that our Token Reconstruction objective could be scaled up for even higher performance and be applied to other domains beyond retinal imaging.

Jesse Phitidis, Alison Q. O'Neil, William N. Whiteley et al.

Cerebrovascular diseases (CVD) can lead to stroke and dementia. Stroke is the second leading cause of death world wide and dementia incidence is increasing by the year. There are several markers of CVD that are visible on brain imaging, including: white matter hyperintensities (WMH), acute and chronic ischaemic stroke lesions (ISL), lacunes, enlarged perivascular spaces (PVS), acute and chronic haemorrhagic lesions, and cerebral microbleeds (CMB). Brain atrophy also occurs in CVD. These markers are important for patient management and intervention, since they indicate elevated risk of future stroke and dementia. We systematically reviewed automated systems designed to support radiologists reporting on these CVD imaging findings. We considered commercially available software and research publications which identify at least two CVD markers. In total, we included 29 commercial products and 13 research publications. Two distinct types of commercial support system were available: those which identify acute stroke lesions (haemorrhagic and ischaemic) from computed tomography (CT) scans, mainly for the purpose of patient triage; and those which measure WMH and atrophy regionally and longitudinally. In research, WMH and ISL were the markers most frequently analysed together, from magnetic resonance imaging (MRI) scans; lacunes and PVS were each targeted only twice and CMB only once. For stroke, commercially available systems largely support the emergency setting, whilst research systems consider also follow-up and routine scans. The systems to quantify WMH and atrophy are focused on neurodegenerative disease support, where these CVD markers are also of significance. There are currently no openly validated systems, commercially, or in research, performing a comprehensive joint analysis of all CVD markers (WMH, ISL, lacunes, PVS, haemorrhagic lesions, CMB, and atrophy).

Ben Philps, Maria del C. Valdes Hernandez, Chen Qin et al.

White Matter Hyperintensities (WMH) are key neuroradiological markers of small vessel disease present in brain MRI. Assessment of WMH is important in research and clinics. However, WMH are challenging to segment due to their high variability in shape, location, size, poorly defined borders, and similar intensity profile to other pathologies (e.g stroke lesions) and artefacts (e.g head motion). In this work, we assess the utility and semantic properties of the most effective techniques for uncertainty quantification (UQ) in segmentation for the WMH segmentation task across multiple test-time data distributions. We find UQ techniques reduce 'silent failure' by identifying in UQ maps small WMH clusters in the deep white matter that are unsegmented by the model. A combination of Stochastic Segmentation Networks with Deep Ensembles also yields the highest Dice and lowest Absolute Volume Difference % (AVD) score and can highlight areas where there is ambiguity between WMH and stroke lesions. We further demonstrate the downstream utility of UQ, proposing a novel method for classification of the clinical Fazekas score using spatial features extracted from voxelwise WMH probability and UQ maps. We show that incorporating WMH uncertainty information improves Fazekas classification performance and calibration. Our model with (UQ and spatial WMH features)/(spatial WMH features)/(WMH volume only) achieves a balanced accuracy score of 0.74/0.67/0.62, and root brier score of 0.65/0.72/0.74 in the Deep WMH and balanced accuracy of 0.74/0.73/0.71 and root brier score of 0.64/0.66/0.68 in the Periventricular region. We further demonstrate that stochastic UQ techniques with high sample diversity can improve the detection of poor quality segmentations.

Justin Engelmann, Amos Storkey, Miguel O. Bernabeu

Deep learning (DL) models are very effective on many computer vision problems and increasingly used in critical applications. They are also inherently black box. A number of methods exist to generate image-wise explanations that allow practitioners to understand and verify model predictions for a given image. Beyond that, it would be desirable to validate that a DL model \textit{generally} works in a sensible way, i.e. consistent with domain knowledge and not relying on undesirable data artefacts. For this purpose, the model needs to be explained globally. In this work, we focus on image modalities that are naturally aligned such that each pixel position represents a similar relative position on the imaged object, as is common in medical imaging. We propose the pixel-wise aggregation of image-wise explanations as a simple method to obtain label-wise and overall global explanations. These can then be used for model validation, knowledge discovery, and as an efficient way to communicate qualitative conclusions drawn from inspecting image-wise explanations. We further propose Progressive Erasing Plus Progressive Restoration (PEPPR) as a method to quantitatively validate that these global explanations are faithful to how the model makes its predictions. We then apply these methods to ultra-widefield retinal images, a naturally aligned modality. We find that the global explanations are consistent with domain knowledge and faithfully reflect the model's workings.

Ylenia Giarratano, Eleonora Bianchi, Calum Gray et al.

Optical coherence tomography angiography (OCTA) is a novel non-invasive imaging modality for the visualisation of microvasculature in vivo that has encountered broad adoption in retinal research. OCTA potential in the assessment of pathological conditions and the reproducibility of studies relies on the quality of the image analysis. However, automated segmentation of parafoveal OCTA images is still an open problem. In this study, we generate the first open dataset of retinal parafoveal OCTA images with associated ground truth manual segmentations. Furthermore, we establish a standard for OCTA image segmentation by surveying a broad range of state-of-the-art vessel enhancement and binarisation procedures. We provide the most comprehensive comparison of these methods under a unified framework to date. Our results show that, for the set of images considered, deep learning architectures (U-Net and CS-Net) achieve the best performance. For applications where manually segmented data is not available to retrain these approaches, our findings suggest that optimal oriented flux is the best handcrafted filter from those considered. Furthermore, we report on the importance of preserving network structure in the segmentation to enable deep vascular phenotyping. We introduce new metrics for network structure evaluation in segmented angiograms. Our results demonstrate that segmentation methods with equal Dice score perform very differently in terms of network structure preservation. Moreover, we compare the error in the computation of clinically relevant vascular network metrics (e.g. foveal avascular zone area and vessel density) across segmentation methods. Our results show up to 25% differences in vessel density accuracy depending on the segmentation method employed. These findings should be taken into account when comparing the results of clinical studies and performing meta-analyses.

Hywel B. Carver, Derek Groen, James Hetherington et al.

We present a new design pattern for high-performance parallel scientific software, named coalesced communication. This pattern allows for a structured way to improve the communication performance through coalescence of multiple communication needs using two communication management components. We apply the design pattern to several simulations of a lattice-Boltzmann blood flow solver with streaming visualisation which engenders a reduction in the communication overhead of approximately 40%.