Ke Zou

Meng Wang, Kai Yu, Chun-Mei Feng et al.

Focusing on the complicated pathological features, such as blurred boundaries, severe scale differences between symptoms, background noise interference, etc., in the task of retinal edema lesions joint segmentation from OCT images and enabling the segmentation results more reliable. In this paper, we propose a novel reliable multi-scale wavelet-enhanced transformer network, which can provide accurate segmentation results with reliability assessment. Specifically, aiming at improving the model's ability to learn the complex pathological features of retinal edema lesions in OCT images, we develop a novel segmentation backbone that integrates a wavelet-enhanced feature extractor network and a multi-scale transformer module of our newly designed. Meanwhile, to make the segmentation results more reliable, a novel uncertainty segmentation head based on the subjective logical evidential theory is introduced to generate the final segmentation results with a corresponding overall uncertainty evaluation score map. We conduct comprehensive experiments on the public database of AI-Challenge 2018 for retinal edema lesions segmentation, and the results show that our proposed method achieves better segmentation accuracy with a high degree of reliability as compared to other state-of-the-art segmentation approaches. The code will be released on: https://github.com/LooKing9218/ReliableRESeg.

Ke Zou, Zhihao Chen, Xuedong Yuan et al.

The use of AI systems in healthcare for the early screening of diseases is of great clinical importance. Deep learning has shown great promise in medical imaging, but the reliability and trustworthiness of AI systems limit their deployment in real clinical scenes, where patient safety is at stake. Uncertainty estimation plays a pivotal role in producing a confidence evaluation along with the prediction of the deep model. This is particularly important in medical imaging, where the uncertainty in the model's predictions can be used to identify areas of concern or to provide additional information to the clinician. In this paper, we review the various types of uncertainty in deep learning, including aleatoric uncertainty and epistemic uncertainty. We further discuss how they can be estimated in medical imaging. More importantly, we review recent advances in deep learning models that incorporate uncertainty estimation in medical imaging. Finally, we discuss the challenges and future directions in uncertainty estimation in deep learning for medical imaging. We hope this review will ignite further interest in the community and provide researchers with an up-to-date reference regarding applications of uncertainty estimation models in medical imaging.

Meng Wang, Tian Lin, Lianyu Wang et al.

Failure to recognize samples from the classes unseen during training is a major limitation of artificial intelligence in the real-world implementation for recognition and classification of retinal anomalies. We established an uncertainty-inspired open-set (UIOS) model, which was trained with fundus images of 9 retinal conditions. Besides assessing the probability of each category, UIOS also calculated an uncertainty score to express its confidence. Our UIOS model with thresholding strategy achieved an F1 score of 99.55%, 97.01% and 91.91% for the internal testing set, external target categories (TC)-JSIEC dataset and TC-unseen testing set, respectively, compared to the F1 score of 92.20%, 80.69% and 64.74% by the standard AI model. Furthermore, UIOS correctly predicted high uncertainty scores, which would prompt the need for a manual check in the datasets of non-target categories retinal diseases, low-quality fundus images, and non-fundus images. UIOS provides a robust method for real-world screening of retinal anomalies.

Ke Zou, Xuedong Yuan, Xiaojing Shen et al.

Despite recent improvements in the accuracy of brain tumor segmentation, the results still exhibit low levels of confidence and robustness. Uncertainty estimation is one effective way to change this situation, as it provides a measure of confidence in the segmentation results. In this paper, we propose a trusted brain tumor segmentation network which can generate robust segmentation results and reliable uncertainty estimations without excessive computational burden and modification of the backbone network. In our method, uncertainty is modeled explicitly using subjective logic theory, which treats the predictions of backbone neural network as subjective opinions by parameterizing the class probabilities of the segmentation as a Dirichlet distribution. Meanwhile, the trusted segmentation framework learns the function that gathers reliable evidence from the feature leading to the final segmentation results. Overall, our unified trusted segmentation framework endows the model with reliability and robustness to out-of-distribution samples. To evaluate the effectiveness of our model in robustness and reliability, qualitative and quantitative experiments are conducted on the BraTS 2019 dataset.

Kai Ren, Ke Zou, Xianjie Liu et al.

Classification and segmentation are crucial in medical image analysis as they enable accurate diagnosis and disease monitoring. However, current methods often prioritize the mutual learning features and shared model parameters, while neglecting the reliability of features and performances. In this paper, we propose a novel Uncertainty-informed Mutual Learning (UML) framework for reliable and interpretable medical image analysis. Our UML introduces reliability to joint classification and segmentation tasks, leveraging mutual learning with uncertainty to improve performance. To achieve this, we first use evidential deep learning to provide image-level and pixel-wise confidences. Then, an Uncertainty Navigator Decoder is constructed for better using mutual features and generating segmentation results. Besides, an Uncertainty Instructor is proposed to screen reliable masks for classification. Overall, UML could produce confidence estimation in features and performance for each link (classification and segmentation). The experiments on the public datasets demonstrate that our UML outperforms existing methods in terms of both accuracy and robustness. Our UML has the potential to explore the development of more reliable and explainable medical image analysis models. We will release the codes for reproduction after acceptance.

Meng Wang, Lianyu Wang, Xinxing Xu et al.

Deep learning models have shown promising performance in the field of diabetic retinopathy (DR) staging. However, collaboratively training a DR staging model across multiple institutions remains a challenge due to non-iid data, client reliability, and confidence evaluation of the prediction. To address these issues, we propose a novel federated uncertainty-aware aggregation paradigm (FedUAA), which considers the reliability of each client and produces a confidence estimation for the DR staging. In our FedUAA, an aggregated encoder is shared by all clients for learning a global representation of fundus images, while a novel temperature-warmed uncertainty head (TWEU) is utilized for each client for local personalized staging criteria. Our TWEU employs an evidential deep layer to produce the uncertainty score with the DR staging results for client reliability evaluation. Furthermore, we developed a novel uncertainty-aware weighting module (UAW) to dynamically adjust the weights of model aggregation based on the uncertainty score distribution of each client. In our experiments, we collect five publicly available datasets from different institutions to conduct a dataset for federated DR staging to satisfy the real non-iid condition. The experimental results demonstrate that our FedUAA achieves better DR staging performance with higher reliability compared to other federated learning methods. Our proposed FedUAA paradigm effectively addresses the challenges of collaboratively training DR staging models across multiple institutions, and provides a robust and reliable solution for the deployment of DR diagnosis models in real-world clinical scenarios.

Meng Wang, Kai Yu, Chun-Mei Feng et al.

Federated learning (FL), as an effective decentralized distributed learning approach, enables multiple institutions to jointly train a model without sharing their local data. However, the domain feature shift caused by different acquisition devices/clients substantially degrades the performance of the FL model. Furthermore, most existing FL approaches aim to improve accuracy without considering reliability (e.g., confidence or uncertainty). The predictions are thus unreliable when deployed in safety-critical applications. Therefore, aiming at improving the performance of FL in non-Domain feature issues while enabling the model more reliable. In this paper, we propose a novel reliable federated disentangling network, termed RFedDis, which utilizes feature disentangling to enable the ability to capture the global domain-invariant cross-client representation and preserve local client-specific feature learning. Meanwhile, to effectively integrate the decoupled features, an uncertainty-aware decision fusion is also introduced to guide the network for dynamically integrating the decoupled features at the evidence level, while producing a reliable prediction with an estimated uncertainty. To the best of our knowledge, our proposed RFedDis is the first work to develop an FL approach based on evidential uncertainty combined with feature disentangling, which enhances the performance and reliability of FL in non-IID domain features. Extensive experimental results show that our proposed RFedDis provides outstanding performance with a high degree of reliability as compared to other state-of-the-art FL approaches.

Ke Zou, Tian Lin, Xuedong Yuan et al.

Multimodality eye disease screening is crucial in ophthalmology as it integrates information from diverse sources to complement their respective performances. However, the existing methods are weak in assessing the reliability of each unimodality, and directly fusing an unreliable modality may cause screening errors. To address this issue, we introduce a novel multimodality evidential fusion pipeline for eye disease screening, EyeMoSt, which provides a measure of confidence for unimodality and elegantly integrates the multimodality information from a multi-distribution fusion perspective. Specifically, our model estimates both local uncertainty for unimodality and global uncertainty for the fusion modality to produce reliable classification results. More importantly, the proposed mixture of Student's $t$ distributions adaptively integrates different modalities to endow the model with heavy-tailed properties, increasing robustness and reliability. Our experimental findings on both public and in-house datasets show that our model is more reliable than current methods. Additionally, EyeMost has the potential ability to serve as a data quality discriminator, enabling reliable decision-making for multimodality eye disease screening.

Guoyao Deng, Ke Zou, Meng Wang et al.

Frontotemporal Dementia (FTD) diagnosis has been successfully progress using deep learning techniques. However, current FTD identification methods suffer from two limitations. Firstly, they do not exploit the potential of multi-view functional magnetic resonance imaging (fMRI) for classifying FTD. Secondly, they do not consider the reliability of the multi-view FTD diagnosis. To address these limitations, we propose a reliable multi-view impartial decision network (MID-Net) for FTD diagnosis in fMRI. Our MID-Net provides confidence for each view and generates a reliable prediction without any conflict. To achieve this, we employ multiple expert models to extract evidence from the abundant neural network information contained in fMRI images. We then introduce the Dirichlet Distribution to characterize the expert class probability distribution from an evidence level. Additionally, a novel Impartial Decision Maker (IDer) is proposed to combine the different opinions inductively to arrive at an unbiased prediction without additional computation cost. Overall, our MID-Net dynamically integrates the decisions of different experts on FTD disease, especially when dealing with multi-view high-conflict cases. Extensive experiments on a high-quality FTD fMRI dataset demonstrate that our model outperforms previous methods and provides high uncertainty for hard-to-classify examples. We believe that our approach represents a significant step toward the deployment of reliable FTD decision-making under multi-expert conditions. We will release the codes for reproduction after acceptance.

Nan Zhou, Ke Zou, Kai Ren et al.

The Medical Segment Anything Model (MedSAM) has shown remarkable performance in medical image segmentation, drawing significant attention in the field. However, its sensitivity to varying prompt types and locations poses challenges. This paper addresses these challenges by focusing on the development of reliable prompts that enhance MedSAM's accuracy. We introduce MedSAM-U, an uncertainty-guided framework designed to automatically refine multi-prompt inputs for more reliable and precise medical image segmentation. Specifically, we first train a Multi-Prompt Adapter integrated with MedSAM, creating MPA-MedSAM, to adapt to diverse multi-prompt inputs. We then employ uncertainty-guided multi-prompt to effectively estimate the uncertainties associated with the prompts and their initial segmentation results. In particular, a novel uncertainty-guided prompts adaptation technique is then applied automatically to derive reliable prompts and their corresponding segmentation outcomes. We validate MedSAM-U using datasets from multiple modalities to train a universal image segmentation model. Compared to MedSAM, experimental results on five distinct modal datasets demonstrate that the proposed MedSAM-U achieves an average performance improvement of 1.7\% to 20.5\% across uncertainty-guided prompts.

Ke Zou, Yidi Chen, Ling Huang et al.

Medical image segmentation is critical for disease diagnosis and treatment assessment. However, concerns regarding the reliability of segmentation regions persist among clinicians, mainly attributed to the absence of confidence assessment, robustness, and calibration to accuracy. To address this, we introduce DEviS, an easily implementable foundational model that seamlessly integrates into various medical image segmentation networks. DEviS not only enhances the calibration and robustness of baseline segmentation accuracy but also provides high-efficiency uncertainty estimation for reliable predictions. By leveraging subjective logic theory, we explicitly model probability and uncertainty for medical image segmentation. Here, the Dirichlet distribution parameterizes the distribution of probabilities for different classes of the segmentation results. To generate calibrated predictions and uncertainty, we develop a trainable calibrated uncertainty penalty. Furthermore, DEviS incorporates an uncertainty-aware filtering module, which designs the metric of uncertainty-calibrated error to filter out-of-distribution data. We conducted validation studies on publicly available datasets, including ISIC2018, KiTS2021, LiTS2017, and BraTS2019, to assess the accuracy and robustness of different backbone segmentation models enhanced by DEviS, as well as the efficiency and reliability of uncertainty estimation.

Yifan Gao, Tao Zhou, Yi Zhou et al.

Medical Visual Grounding (MVG) aims to identify diagnostically relevant phrases from free-text radiology reports and localize their corresponding regions in medical images, providing interpretable visual evidence to support clinical decision-making. Although recent Vision-Language Models (VLMs) exhibit promising multimodal reasoning ability, their grounding remains insufficient spatial precision, largely due to a lack of explicit localization priors when relying solely on latent embeddings. In this work, we analyze this limitation from an attention perspective and propose KnowMVG, a Knowledge-prior and global-local attention enhancement framework for MVG in VLMs that explicitly strengthens spatial awareness during decoding. Specifically, we present a knowledge-enhanced prompting strategy that encodes phrase related medical knowledge into compact embeddings, together with a global-local attention that jointly leverages coarse global information and refined local cues to guide precise region localization. localization. This design bridges high-level semantic understanding and fine-grained visual perception without introducing extra textual reasoning overhead. Extensive experiments on four MVG benchmarks demonstrate that our KnowMVG consistently outperforms existing approaches, achieving gains of 3.0% in AP50 and 2.6% in mIoU over prior state-of-the-art methods. Qualitative and ablation studies further validate the effectiveness of each component.

Xiaoyu Pan, Yang Bai, Ke Zou et al.

Medical Large Language Models (MLLMs) play a crucial role in ophthalmic diagnosis, holding significant potential to address vision-threatening diseases. However, their accuracy is constrained by hallucinations stemming from limited ophthalmic knowledge, insufficient visual localization and reasoning capabilities, and a scarcity of multimodal ophthalmic data, which collectively impede precise lesion detection and disease diagnosis. Furthermore, existing medical benchmarks fail to effectively evaluate various types of hallucinations or provide actionable solutions to mitigate them. To address the above challenges, we introduce EH-Benchmark, a novel ophthalmology benchmark designed to evaluate hallucinations in MLLMs. We categorize MLLMs' hallucinations based on specific tasks and error types into two primary classes: Visual Understanding and Logical Composition, each comprising multiple subclasses. Given that MLLMs predominantly rely on language-based reasoning rather than visual processing, we propose an agent-centric, three-phase framework, including the Knowledge-Level Retrieval stage, the Task-Level Case Studies stage, and the Result-Level Validation stage. Experimental results show that our multi-agent framework significantly mitigates both types of hallucinations, enhancing accuracy, interpretability, and reliability. Our project is available at https://github.com/ppxy1/EH-Benchmark.

Qingshan Hou, Yukun Zhou, Jocelyn Hui Lin Goh et al.

The advent of foundation models (FMs) is transforming medical domain. In ophthalmology, RETFound, a retina-specific FM pre-trained sequentially on 1.4 million natural images and 1.6 million retinal images, has demonstrated high adaptability across clinical applications. Conversely, DINOv2, a general-purpose vision FM pre-trained on 142 million natural images, has shown promise in non-medical domains. However, its applicability to clinical tasks remains underexplored. To address this, we conducted head-to-head evaluations by fine-tuning RETFound and three DINOv2 models (large, base, small) for ocular disease detection and systemic disease prediction tasks, across eight standardized open-source ocular datasets, as well as the Moorfields AlzEye and the UK Biobank datasets. DINOv2-large model outperformed RETFound in detecting diabetic retinopathy (AUROC=0.850-0.952 vs 0.823-0.944, across three datasets, all P<=0.007) and multi-class eye diseases (AUROC=0.892 vs. 0.846, P<0.001). In glaucoma, DINOv2-base model outperformed RETFound (AUROC=0.958 vs 0.940, P<0.001). Conversely, RETFound achieved superior performance over all DINOv2 models in predicting heart failure, myocardial infarction, and ischaemic stroke (AUROC=0.732-0.796 vs 0.663-0.771, all P<0.001). These trends persisted even with 10% of the fine-tuning data. These findings showcase the distinct scenarios where general-purpose and domain-specific FMs excel, highlighting the importance of aligning FM selection with task-specific requirements to optimise clinical performance.

Samantha Min Er Yew, Xiaofeng Lei, Jocelyn Hui Lin Goh et al.

Background: RETFound, a self-supervised, retina-specific foundation model (FM), showed potential in downstream applications. However, its comparative performance with traditional deep learning (DL) models remains incompletely understood. This study aimed to evaluate RETFound against three ImageNet-pretrained supervised DL models (ResNet50, ViT-base, SwinV2) in detecting ocular and systemic diseases. Methods: We fine-tuned/trained RETFound and three DL models on full datasets, 50%, 20%, and fixed sample sizes (400, 200, 100 images, with half comprising disease cases; for each DR severity class, 100 and 50 cases were used. Fine-tuned models were tested internally using the SEED (53,090 images) and APTOS-2019 (3,672 images) datasets and externally validated on population-based (BES, CIEMS, SP2, UKBB) and open-source datasets (ODIR-5k, PAPILA, GAMMA, IDRiD, MESSIDOR-2). Model performance was compared using area under the receiver operating characteristic curve (AUC) and Z-tests with Bonferroni correction (P<0.05/3). Interpretation: Traditional DL models are mostly comparable to RETFound for ocular disease detection with large datasets. However, RETFound is superior in systemic disease detection with smaller datasets. These findings offer valuable insights into the respective merits and limitation of traditional models and FMs.

Bo Liu, Ke Zou, Liming Zhan et al.

Medical Visual Question Answering (Med-VQA) combines computer vision and natural language processing to automatically answer clinical inquiries about medical images. However, current Med-VQA datasets exhibit two significant limitations: (1) they often lack visual and textual explanations for answers, hindering comprehension for patients and junior doctors; (2) they typically offer a narrow range of question formats, inadequately reflecting the diverse requirements in practical scenarios. These limitations pose significant challenges to the development of a reliable and user-friendly Med-VQA system. To address these challenges, we introduce a large-scale, Groundable, and Explainable Medical VQA benchmark for chest X-ray diagnosis (GEMeX), featuring several innovative components: (1) a multi-modal explainability mechanism that offers detailed visual and textual explanations for each question-answer pair, thereby enhancing answer comprehensibility; (2) four question types, open-ended, closed-ended, single-choice, and multiple-choice, to better reflect practical needs. With 151,025 images and 1,605,575 questions, GEMeX is the currently largest chest X-ray VQA dataset. Evaluation of 12 representative large vision language models (LVLMs) on GEMeX reveals suboptimal performance, underscoring the dataset's complexity. Meanwhile, we propose a strong model by fine-tuning an existing LVLM on the GEMeX training set. The substantial performance improvement showcases the dataset's effectiveness. The benchmark is available at https://www.med-vqa.com/GEMeX.

Ke Zou, Yang Bai, Bo Liu et al.

Medical phrase grounding is crucial for identifying relevant regions in medical images based on phrase queries, facilitating accurate image analysis and diagnosis. However, current methods rely on manual extraction of key phrases from medical reports, reducing efficiency and increasing the workload for clinicians. Additionally, the lack of model confidence estimation limits clinical trust and usability. In this paper, we introduce a novel task called Medical Report Grounding (MRG), which aims to directly identify diagnostic phrases and their corresponding grounding boxes from medical reports in an end-to-end manner. To address this challenge, we propose uMedGround, a robust and reliable framework that leverages a multimodal large language model to predict diagnostic phrases by embedding a unique token, <BOX>, into the vocabulary to enhance detection capabilities. A vision encoder-decoder processes the embedded token and input image to generate grounding boxes. Critically, uMedGround incorporates an uncertainty-aware prediction model, significantly improving the robustness and reliability of grounding predictions. Experimental results demonstrate that uMedGround outperforms state-of-the-art medical phrase grounding methods and fine-tuned large visual-language models, validating its effectiveness and reliability. This study represents a pioneering exploration of the MRG task, marking the first-ever endeavor in this domain. Additionally, we demonstrate the applicability of uMedGround in medical visual question answering and class-based localization tasks, where it highlights visual evidence aligned with key diagnostic phrases, supporting clinicians in interpreting various types of textual inputs, including free-text reports, visual question answering queries, and class labels.

Zehui Liao, Shishuai Hu, Ke Zou et al.

Multimodal large language models (MLLMs) have demonstrated significant potential in medical Visual Question Answering (VQA). Yet, they remain prone to hallucinations-incorrect responses that contradict input images, posing substantial risks in clinical decision-making. Detecting these hallucinations is essential for establishing trust in MLLMs among clinicians and patients, thereby enabling their real-world adoption. Current hallucination detection methods, especially semantic entropy (SE), have demonstrated promising hallucination detection capacity for LLMs. However, adapting SE to medical MLLMs by incorporating visual perturbations presents a dilemma. Weak perturbations preserve image content and ensure clinical validity, but may be overlooked by medical MLLMs, which tend to over rely on language priors. In contrast, strong perturbations can distort essential diagnostic features, compromising clinical interpretation. To address this issue, we propose Vision Amplified Semantic Entropy (VASE), which incorporates weak image transformations and amplifies the impact of visual input, to improve hallucination detection in medical VQA. We first estimate the semantic predictive distribution under weak visual transformations to preserve clinical validity, and then amplify visual influence by contrasting this distribution with that derived from a distorted image. The entropy of the resulting distribution is estimated as VASE. Experiments on two medical open-ended VQA datasets demonstrate that VASE consistently outperforms existing hallucination detection methods.

Sahana Srinivasan, Xuguang Ai, Minjie Zou et al.

Question: What is the performance and reasoning ability of OpenAI o1 compared to other large language models in addressing ophthalmology-specific questions? Findings: This study evaluated OpenAI o1 and five LLMs using 6,990 ophthalmological questions from MedMCQA. O1 achieved the highest accuracy (0.88) and macro-F1 score but ranked third in reasoning capabilities based on text-generation metrics. Across subtopics, o1 ranked first in ``Lens'' and ``Glaucoma'' but second to GPT-4o in ``Corneal and External Diseases'', ``Vitreous and Retina'' and ``Oculoplastic and Orbital Diseases''. Subgroup analyses showed o1 performed better on queries with longer ground truth explanations. Meaning: O1's reasoning enhancements may not fully extend to ophthalmology, underscoring the need for domain-specific refinements to optimize performance in specialized fields like ophthalmology.

Minjie Zou, Sahana Srinivasan, Thaddaeus Wai Soon Lo et al.

Recent advances in reasoning-focused large language models (LLMs) mark a shift from general LLMs toward models designed for complex decision-making, a crucial aspect in medicine. However, their performance in specialized domains like ophthalmology remains underexplored. This study comprehensively evaluated and compared the accuracy and reasoning capabilities of four newly developed reasoning-focused LLMs, namely DeepSeek-R1, OpenAI o1, o3-mini, and Gemini 2.0 Flash-Thinking. Each model was assessed using 5,888 multiple-choice ophthalmology exam questions from the MedMCQA dataset in zero-shot setting. Quantitative evaluation included accuracy, Macro-F1, and five text-generation metrics (ROUGE-L, METEOR, BERTScore, BARTScore, and AlignScore), computed against ground-truth reasonings. Average inference time was recorded for a subset of 100 randomly selected questions. Additionally, two board-certified ophthalmologists qualitatively assessed clarity, completeness, and reasoning structure of responses to differential diagnosis questions.O1 (0.902) and DeepSeek-R1 (0.888) achieved the highest accuracy, with o1 also leading in Macro-F1 (0.900). The performance of models across the text-generation metrics varied: O3-mini excelled in ROUGE-L (0.151), o1 in METEOR (0.232), DeepSeek-R1 and o3-mini tied for BERTScore (0.673), DeepSeek-R1 (-4.105) and Gemini 2.0 Flash-Thinking (-4.127) performed best in BARTScore, while o3-mini (0.181) and o1 (0.176) led AlignScore. Inference time across the models varied, with DeepSeek-R1 being slowest (40.4 seconds) and Gemini 2.0 Flash-Thinking fastest (6.7 seconds). Qualitative evaluation revealed that DeepSeek-R1 and Gemini 2.0 Flash-Thinking tended to provide detailed and comprehensive intermediate reasoning, whereas o1 and o3-mini displayed concise and summarized justifications.

Ke Zou, Jocelyn Hui Lin Goh, Yukun Zhou et al.

Foundation models (FMs) have shown great promise in medical image analysis by improving generalization across diverse downstream tasks. In ophthalmology, several FMs have recently emerged, but there is still no clear answer to fundamental questions: Which FM performs the best? Are they equally good across different tasks? What if we combine all FMs together? To our knowledge, this is the first study to systematically evaluate both single and fused ophthalmic FMs. To address these questions, we propose FusionFM, a comprehensive evaluation suite, along with two fusion approaches to integrate different ophthalmic FMs. Our framework covers both ophthalmic disease detection (glaucoma, diabetic retinopathy, and age-related macular degeneration) and systemic disease prediction (diabetes and hypertension) based on retinal imaging. We benchmarked four state-of-the-art FMs (RETFound, VisionFM, RetiZero, and DINORET) using standardized datasets from multiple countries and evaluated their performance using AUC and F1 metrics. Our results show that DINORET and RetiZero achieve superior performance in both ophthalmic and systemic disease tasks, with RetiZero exhibiting stronger generalization on external datasets. Regarding fusion strategies, the Gating-based approach provides modest improvements in predicting glaucoma, AMD, and hypertension. Despite these advances, predicting systemic diseases, especially hypertension in external cohort remains challenging. These findings provide an evidence-based evaluation of ophthalmic FMs, highlight the benefits of model fusion, and point to strategies for enhancing their clinical applicability.

Meng Wang, Tian Lin, Qingshan Hou et al.

Artificial intelligence (AI) shows remarkable potential in medical imaging diagnostics, yet most current models require retraining when applied across different clinical settings, limiting their scalability. We introduce GlobeReady, a clinician-friendly AI platform that enables fundus disease diagnosis that operates without retraining, fine-tuning, or the needs for technical expertise. GlobeReady demonstrates high accuracy across imaging modalities: 93.9-98.5% for 11 fundus diseases using color fundus photographs (CPFs) and 87.2-92.7% for 15 fundus diseases using optic coherence tomography (OCT) scans. By leveraging training-free local feature augmentation, GlobeReady platform effectively mitigates domain shifts across centers and populations, achieving accuracies of 88.9-97.4% across five centers on average in China, 86.3-96.9% in Vietnam, and 73.4-91.0% in Singapore, and 90.2-98.9% in the UK. Incorporating a bulit-in confidence-quantifiable diagnostic mechanism further enhances the platform's accuracy to 94.9-99.4% with CFPs and 88.2-96.2% with OCT, while enabling identification of out-of-distribution cases with 86.3% accuracy across 49 common and rare fundus diseases using CFPs, and 90.6% accuracy across 13 diseases using OCT. Clinicians from countries rated GlobeReady highly for usability and clinical relevance (average score 4.6/5). These findings demonstrate GlobeReady's robustness, generalizability and potential to support global ophthalmic care without technical barriers.

Zhenyue Qin, Yang Liu, Yu Yin et al.

Vision-threatening eye diseases pose a major global health burden, with timely diagnosis limited by workforce shortages and restricted access to specialized care. While multimodal large language models (MLLMs) show promise for medical image interpretation, advancing MLLMs for ophthalmology is hindered by the lack of comprehensive benchmark datasets suitable for evaluating generative models. We present a large-scale multimodal ophthalmology benchmark comprising 32,633 instances with multi-granular annotations across 12 common ophthalmic conditions and 5 imaging modalities. The dataset integrates imaging, anatomical structures, demographics, and free-text annotations, supporting anatomical structure recognition, disease screening, disease staging, and demographic prediction for bias evaluation. This work extends our preliminary LMOD benchmark with three major enhancements: (1) nearly 50% dataset expansion with substantial enlargement of color fundus photography; (2) broadened task coverage including binary disease diagnosis, multi-class diagnosis, severity classification with international grading standards, and demographic prediction; and (3) systematic evaluation of 24 state-of-the-art MLLMs. Our evaluations reveal both promise and limitations. Top-performing models achieved ~58% accuracy in disease screening under zero-shot settings, and performance remained suboptimal for challenging tasks like disease staging. We will publicly release the dataset, curation pipeline, and leaderboard to potentially advance ophthalmic AI applications and reduce the global burden of vision-threatening diseases.

Yukun Zhou, Paul Nderitu, Jocelyn Hui Lin Goh et al.

Medical foundation models, pre-trained with large-scale clinical data, demonstrate strong performance in diverse clinically relevant applications. RETFound, trained on nearly one million retinal images, exemplifies this approach in applications with retinal images. However, the emergence of increasingly powerful and multifold larger generalist foundation models such as DINOv2 and DINOv3 raises the question of whether domain-specific pre-training remains essential, and if so, what gap persists. To investigate this, we systematically evaluated the adaptability of DINOv2 and DINOv3 in retinal image applications, compared to two specialist RETFound models, RETFound-MAE and RETFound-DINOv2. We assessed performance on ocular disease detection and systemic disease prediction using two adaptation strategies: fine-tuning and linear probing. Data efficiency and adaptation efficiency were further analysed to characterise trade-offs between predictive performance and computational cost. Our results show that although scaling generalist models yields strong adaptability across diverse tasks, RETFound-DINOv2 consistently outperforms these generalist foundation models in ocular-disease detection and oculomics tasks, demonstrating stronger generalisability and data efficiency. These findings suggest that specialist retinal foundation models remain the most effective choice for clinical applications, while the narrowing gap with generalist foundation models suggests that continued data and model scaling can deliver domain-relevant gains and position them as strong foundations for future medical foundation models.

Sahana Srinivasan, Xuguang Ai, Thaddaeus Wai Soon Lo et al.

Current benchmarks evaluating large language models (LLMs) in ophthalmology are limited in scope and disproportionately prioritise accuracy. We introduce BELO (BEnchmarking LLMs for Ophthalmology), a standardized and comprehensive evaluation benchmark developed through multiple rounds of expert checking by 13 ophthalmologists. BELO assesses ophthalmology-related clinical accuracy and reasoning quality. Using keyword matching and a fine-tuned PubMedBERT model, we curated ophthalmology-specific multiple-choice-questions (MCQs) from diverse medical datasets (BCSC, MedMCQA, MedQA, BioASQ, and PubMedQA). The dataset underwent multiple rounds of expert checking. Duplicate and substandard questions were systematically removed. Ten ophthalmologists refined the explanations of each MCQ's correct answer. This was further adjudicated by three senior ophthalmologists. To illustrate BELO's utility, we evaluated six LLMs (OpenAI o1, o3-mini, GPT-4o, DeepSeek-R1, Llama-3-8B, and Gemini 1.5 Pro) using accuracy, macro-F1, and five text-generation metrics (ROUGE-L, BERTScore, BARTScore, METEOR, and AlignScore). In a further evaluation involving human experts, two ophthalmologists qualitatively reviewed 50 randomly selected outputs for accuracy, comprehensiveness, and completeness. BELO consists of 900 high-quality, expert-reviewed questions aggregated from five sources: BCSC (260), BioASQ (10), MedMCQA (572), MedQA (40), and PubMedQA (18). A public leaderboard has been established to promote transparent evaluation and reporting. Importantly, the BELO dataset will remain a hold-out, evaluation-only benchmark to ensure fair and reproducible comparisons of future models.

Yuanyuan Peng, Aidi Lin, Meng Wang et al.

Inability to express the confidence level and detect unseen classes has limited the clinical implementation of artificial intelligence in the real-world. We developed a foundation model with uncertainty estimation (FMUE) to detect 11 retinal conditions on optical coherence tomography (OCT). In the internal test set, FMUE achieved a higher F1 score of 96.76% than two state-of-the-art algorithms, RETFound and UIOS, and got further improvement with thresholding strategy to 98.44%. In the external test sets obtained from other OCT devices, FMUE achieved an accuracy of 88.75% and 92.73% before and after thresholding. Our model is superior to two ophthalmologists with a higher F1 score (95.17% vs. 61.93% &71.72%). Besides, our model correctly predicts high uncertainty scores for samples with ambiguous features, of non-target-category diseases, or with low-quality to prompt manual checks and prevent misdiagnosis. FMUE provides a trustworthy method for automatic retinal anomalies detection in the real-world clinical open set environment.

Meng Wang, Tian Lin, Aidi Lin et al.

Previous foundation models for fundus images were pre-trained with limited disease categories and knowledge base. Here we introduce a knowledge-rich vision-language model (RetiZero) that leverages knowledge from more than 400 fundus diseases. For RetiZero's pretraining, we compiled 341,896 fundus images paired with texts, sourced from public datasets, ophthalmic literature, and online resources, encompassing a diverse range of diseases across multiple ethnicities and countries. RetiZero exhibits remarkable performance in several downstream tasks, including zero-shot disease recognition, image-to-image retrieval, AI-assisted clinical diagnosis,few-shot fine-tuning, and internal- and cross-domain disease identification. In zero-shot scenarios, RetiZero achieves Top-5 accuracies of 0.843 for 15 diseases and 0.756 for 52 diseases. For image retrieval, it achieves Top-5 scores of 0.950 and 0.886 for the same sets, respectively. AI-assisted clinical diagnosis results show that RetiZero's Top-3 zero-shot performance surpasses the average of 19 ophthalmologists from Singapore, China, and the United States. RetiZero substantially enhances clinicians' accuracy in diagnosing fundus diseases, in particularly rare ones. These findings underscore the value of integrating the RetiZero into clinical settings, where various fundus diseases are encountered.

Hongye Zeng, Ke Zou, Zhihao Chen et al.

Handheld ultrasound devices face usage limitations due to user inexperience and cannot benefit from supervised deep learning without extensive expert annotations. Moreover, the models trained on standard ultrasound device data are constrained by training data distribution and perform poorly when directly applied to handheld device data. In this study, we propose the Training-free Image Style Alignment (TISA) framework to align the style of handheld device data to those of standard devices. The proposed TISA can directly infer handheld device images without extra training and is suited for clinical applications. We show that TISA performs better and more stably in medical detection and segmentation tasks for handheld device data. We further validate TISA as the clinical model for automatic measurements of spinal curvature and carotid intima-media thickness. The automatic measurements agree well with manual measurements made by human experts and the measurement errors remain within clinically acceptable ranges. We demonstrate the potential for TISA to facilitate automatic diagnosis on handheld ultrasound devices and expedite their eventual widespread use.

Guoyao Deng, Ke Zou, Kai Ren et al.

Recently, Segmenting Anything has taken an important step towards general artificial intelligence. At the same time, its reliability and fairness have also attracted great attention, especially in the field of health care. In this study, we propose multi-box prompts triggered uncertainty estimation for SAM cues to demonstrate the reliability of segmented lesions or tissues. We estimate the distribution of SAM predictions via Monte Carlo with prior distribution parameters, which employs different prompts as formulation of test-time augmentation. Our experimental results found that multi-box prompts augmentation improve the SAM performance, and endowed each pixel with uncertainty. This provides the first paradigm for a reliable SAM.