Pearse A Keane

Ehsan Misaghi, Sean T Berkowitz, Bing Yu Chen et al.

Large language models (LLMs) show potential for ophthalmic clinical reasoning, yet individual models risk introducing harm. We evaluated whether multi-agent LLM deliberative councils improve diagnostic performance and mitigate harm compared to individual LLMs. In a comparative cross-sectional study, we assessed 12 individual LLMs and three multi-agent councils on 100 ophthalmology clinical vignettes. Each council comprised four models assembled by type: proprietary flagship, proprietary fast, and open-source. Models independently answered a vignette, anonymously ranked one another's responses, and a designated chair synthesized all responses and peer reviews into a final answer. Councils consistently outperformed pooled individual models across all three tiers. Accuracy improved for proprietary flagship (95.0% vs 90.8%; risk difference [RD]: 4.25 [95% CI: 0.45, 8.05]), proprietary fast (96.0% vs 86.5%; RD: 9.50 [5.31, 13.59]), and open-source councils (91.0% vs 83.2%; RD: 7.75 [4.17, 11.33]). Harm rates declined for proprietary flagship (10.0% vs 22.5%; RD: -12.50 [-16.86, -8.14]), proprietary fast (16.0% vs 31.8%; RD: -15.75 [-21.49, -10.01]), and open-source councils (22.0% vs 38.5%; RD: -16.50 [-22.27, -10.73]). Coverage analysis revealed net positive gains for accuracy (ΔCoverage: 4.4-9.8 percentage points) and safety (ΔCoverage: 13.6-20.6), indicating councils recovered correct diagnoses and averted harm. Councils elevated correct diagnoses to higher rank positions; and produced more complete differentials and management plans (all P<.05). Harmful council responses showed reduced combined commission-and-omission errors and tended to be less severe. Structured deliberation via multi-agent LLM councils may enhance the reliability of LLM-assisted ophthalmic clinical reasoning.

Ariel Yuhan Ong, Iain Livingstone, Caroline Kilduff et al.

Clinicians often face workflow problems that are perceived as either too bespoke or low stakes to attract commercial attention. Historically, most do not have the technical knowledge to address these problems, but the recent emergence of "vibe coding" presents a transformative opportunity. Vibe coding refers to the co-development of software using natural language prompts to large language models. It offers a pathway to create simple tools that address these real-world pain points, or to prototype more complex ideas. In this review, written by a group of early adopter clinicians with a range of programming expertise, we introduce vibe coding for clinicians (especially those with no or minimal coding experience) as a way of democratising innovation from the front lines. We discuss foundational skills, outline some common challenges, provide a practical step-by-step playbook, and illustrate this approach with some case examples, taking care to consider caveats and guardrails for deployment. We propose that vibe coding is more than a technical shortcut for beginners and is not a replacement for professional software developers. Instead, it can bridge the gap between clinical insight and technical execution, equipping clinicians with the ability to rapidly prototype digital health solutions most reflective of clinical realities.

Fares Antaki, David Mikhail, Daniel Milad et al.

Large language models (LLMs) such as GPT-5 integrate advanced reasoning capabilities that may improve performance on complex medical question-answering tasks. For this latest generation of reasoning models, the configurations that maximize both accuracy and cost-efficiency have yet to be established. We evaluated 12 configurations of OpenAI's GPT-5 series (three model tiers across four reasoning effort settings) alongside o1-high, o3-high, and GPT-4o, using 260 closed-access multiple-choice questions from the American Academy of Ophthalmology Basic Clinical Science Course (BCSC) dataset. The primary outcome was multiple-choice accuracy; secondary outcomes included head-to-head ranking via a Bradley-Terry model, rationale quality assessment using a reference-anchored, pairwise LLM-as-a-judge framework, and analysis of accuracy-cost trade-offs using token-based cost estimates. GPT-5-high achieved the highest accuracy (0.965; 95% CI, 0.942-0.985), outperforming all GPT-5-nano variants (P < .001), o1-high (P = .04), and GPT-4o (P < .001), but not o3-high (0.958; 95% CI, 0.931-0.981). GPT-5-high ranked first in both accuracy (1.66x stronger than o3-high) and rationale quality (1.11x stronger than o3-high). Cost-accuracy analysis identified several GPT-5 configurations on the Pareto frontier, with GPT-5-mini-low offering the most favorable low-cost, high-performance balance. These results benchmark GPT-5 on a high-quality ophthalmology dataset, demonstrate the influence of reasoning effort on accuracy, and introduce an autograder framework for scalable evaluation of LLM-generated answers against reference standards in ophthalmology.

Zijie Cheng, Boxuan Li, André Altmann et al.

Contrastive learning, a prominent approach within self-supervised learning, has demonstrated significant effectiveness in developing generalizable models for various applications involving natural images. However, recent research indicates that these successes do not necessarily extend to the medical imaging domain. In this paper, we investigate the reasons for this suboptimal performance and hypothesize that the dense distribution of medical images poses challenges to the pretext tasks in contrastive learning, particularly in constructing positive and negative pairs. We explore model performance under different augmentation strategies and compare the results to those achieved with strong augmentations. Our study includes six publicly available datasets covering multiple clinically relevant tasks. We further assess the model's generalizability through external evaluations. The model pre-trained with weak augmentation outperforms those with strong augmentation, improving AUROC from 0.838 to 0.848 and AUPR from 0.523 to 0.597 on MESSIDOR2, and showing similar enhancements across other datasets. Our findings suggest that optimizing the scale of augmentation is critical for enhancing the efficacy of contrastive learning in medical imaging.

Avinash Varadarajan, Pinal Bavishi, Paisan Raumviboonsuk et al.

Diabetic eye disease is one of the fastest growing causes of preventable blindness. With the advent of anti-VEGF (vascular endothelial growth factor) therapies, it has become increasingly important to detect center-involved diabetic macular edema (ci-DME). However, center-involved diabetic macular edema is diagnosed using optical coherence tomography (OCT), which is not generally available at screening sites because of cost and workflow constraints. Instead, screening programs rely on the detection of hard exudates in color fundus photographs as a proxy for DME, often resulting in high false positive or false negative calls. To improve the accuracy of DME screening, we trained a deep learning model to use color fundus photographs to predict ci-DME. Our model had an ROC-AUC of 0.89 (95% CI: 0.87-0.91), which corresponds to a sensitivity of 85% at a specificity of 80%. In comparison, three retinal specialists had similar sensitivities (82-85%), but only half the specificity (45-50%, p<0.001 for each comparison with model). The positive predictive value (PPV) of the model was 61% (95% CI: 56-66%), approximately double the 36-38% by the retinal specialists. In addition to predicting ci-DME, our model was able to detect the presence of intraretinal fluid with an AUC of 0.81 (95% CI: 0.81-0.86) and subretinal fluid with an AUC of 0.88 (95% CI: 0.85-0.91). The ability of deep learning algorithms to make clinically relevant predictions that generally require sophisticated 3D-imaging equipment from simple 2D images has broad relevance to many other applications in medical imaging.