Hasil untuk "Ophthalmology"

Zhenyue Qin, Younjoon Chung, Elijah Lee et al.

Vision impairment affects millions globally, and early detection is critical to preventing irreversible vision loss. Ophthalmology workflows require clinicians to integrate medical images, structured clinical data, and free-text notes to determine disease severity and management, which is time-consuming and burdensome. Recent multimodal large language models (MLLMs) show promise, but existing general and medical MLLMs perform poorly in ophthalmology, and few ophthalmology-specific MLLMs are openly available. We present VOLMO (Versatile and Open Large Models for Ophthalmology), a model-agnostic, data-open framework for developing ophthalmology-specific MLLMs. VOLMO includes three stages: ophthalmology knowledge pretraining on 86,965 image-text pairs from 26,569 articles across 82 journals; domain task fine-tuning on 26,929 annotated instances spanning 12 eye conditions for disease screening and severity classification; and multi-step clinical reasoning on 913 patient case reports for assessment, planning, and follow-up care. Using this framework, we trained a compact 2B-parameter MLLM and compared it with strong baselines, including InternVL-2B, LLaVA-Med-7B, MedGemma-4B, MedGemma-27B, and RETFound. We evaluated these models on image description generation, disease screening and staging classification, and assessment-and-management generation, with additional manual review by two healthcare professionals and external validation on three independent cohorts for age-related macular degeneration and diabetic retinopathy. Across settings, VOLMO-2B consistently outperformed baselines, achieving stronger image description performance, an average F1 of 87.4% across 12 eye conditions, and higher scores in external validation.

Jinyoung Moon, Yongseok Mun

Abstract Background The US National Health and Nutrition Examination Survey (NHANES) dataset does not include a specific question or laboratory test to confirm a history of cancer diagnosis. However, if straightforward variables for cancer history are introduced, US NHANES could be effectively utilized in future cancer epidemiology studies. To address this gap, the authors developed a cancer patient database from the US NHANES datasets by employing multiple R programming codes. Methods To illustrate the practical application of this methodology to a real-world problem, the authors extracted the R codes applied in an academic paper published in another journal on January 30th, 2024 ( https://doi.org/10.1016/j.heliyon.2024.e24337 ). This paper will focus on the construction of the database and analysis using R codes. Entire. Results In the first example, the urine concentration of monocarboxynonyl phthalate, monocarboxyoctyl phthalate, mono-2-ethyl-5-carboxypentyl phthalate, and mono-2-hydroxy-iso-butyl phthalate (all ng/mL) were used as the independent variable, instead of the serum concentration of perfluorooctanoic acid (PFOA), perfluorooctane sulfonic acid (PFOS), perfluorohexane sulfonic acid (PFHxS), and perfluorononanoic acid (PFNA), respectively. In the second example, the serum concentration of 2,3,3’,4,4’-Pentachlorobiphenyl (PCB105), 2,3,4,4´,5-Pentachlorobiphenyl (PCB114), 2,3’,4,4’,5-Pentachlorobiphenyl (PCB118), and 2,2’,3,4,4’,5’- and 2,3,3’,4,4’,6-Hexachlorobiphenyl (PCB138) were used as the independent variable, instead of the serum concentration of PFOA, PFOS, PFHxS, and PFNA, respectively. Discussion This research offers a comprehensive set of R codes aimed at creating a single, user-friendly variable that encapsulates the history of each type of cancer while also considering the age at which the diagnosis was made. The US NHANES provides a wealth of critical data on environmental toxicant exposures. By employing these R codes, researchers can potentially discover numerous new associations between environmental toxicant exposures and cancer diagnoses. Ultimately, these codes could significantly advance the field of cancer epidemiology in relation to environmental toxicant exposure.

Rajneesh Dhiman, Barkha Gupta, Arun Kumar Jain et al.

Purpose: To compare topographic, pachymetric and epithelial distribution patterns of clinically unaffected fellow eyes of patients with asymmetric keratoconus (AKC) with those of normal controls and keratoconic fellow eyes. Methods: A prospective, observational study was conducted at a tertiary care center, in which twenty-five clinically unaffected fellow eyes in patients with AKC were compared to 25 keratoconic fellow eyes and 67 normal controls. Pachymetric and epithelial thicknesses in the central 2 mm and 8 octants in the 2–5 mm and 5–7 mm zones were compared using Spectral Domain Optical Coherence Tomography (SDOCT). Scheimpflug derived topography was also compared. Results: Mean simulated keratometry of fellow eyes (43.6D ± 1.34 D) was comparable to that of control eyes (43.85D ± 1.57 D) (p > 0.99) and less than that of keratoconic eyes (47.08D ± 4.62 D) (p = 0.004). Central corneal and epithelial thicknesses of fellow eyes (474.28 ± 40.27 μm, 52.76 ± 7.45 μm, respectively) were less than in control eyes (505.97 ± 30.12 μm, 60.48 ± 8.37 μm) (p < 0.001 and p = 0.006, respectively) but comparable to those of keratoconic eyes (470.48 ± 44.19 μm, 52.48 ± 8.00 μm) (p > 0.99). Pachymetric differences of radially opposite octants in the 2–5 mm zone demonstrated thinning in the inferior and inferotemporal octants of fellow eyes which was greater than that seen in control eyes (p = 0.017 and 0.04, respectively) Conclusion: Reduced central epithelial thickness and greater pachymetric differences in radially opposite octants of the mid-peripheral cornea may be suggestive of susceptibility to ectasia despite normal topography.

Eliza Williams, Jeremy Altman, Saleh Ahmed et al.

Abstract Background Alterations in the constituents of the aqueous humor (AH) are associated with various ocular pathologies, including primary open-angle glaucoma (POAG). AH contains a variety of immunomodulatory molecules, including serine protease inhibitors (serpins), which regulate several proteolytic cascades such as coagulation, angiogenesis, and inflammation. The purpose of this study was to examine the levels of different serpins in human AH and their association with POAG pathology. Methods The abundance of all 37 serpins was determined using LC-MS/MS analysis in 289 human AH samples (cataract: n = 209; POAG: n = 80). The potential involvement of these serpins in POAG was examined by correlating their levels with clinical parameters such as intraocular pressure (IOP) and optic nerve damage. Results Among the 37 serpins present in the human proteome, 26 were detected in aqueous humor. The thirteen most abundant serpins in AH include SERPINA1, SERPINF1, SERPINC1, SERPINA3, SERPING1, AGT, SERPINF2, SERPINA4, SERPINA6, SERPIND1, SERPINI1, SERPINA7, and SERPINA5. Seven serpins were downregulated in subjects with POAG, including SERPINI1 (FC = 0.26), SERPINA4 (FC = 0.40), SERPINA6 (FC = 0.42), SERPINA7 (FC = 0.46), SERPINC1 (FC = 0.74), AGT (FC = 0.76), and SERPING1 (FC = 0.78). Conclusion This study highlights significant alterations in serpin levels within the AH of individuals with POAG. Sex-specific and race-specific differences in the levels of several serpins were also observed. Further studies are needed to clarify the specific mechanisms by which these serpins may contribute to POAG progression and to investigate their potential clinical relevance.

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.

Chengfeng Zhou, Ji Wang, Juanjuan Qin et al.

Large language models (LLMs) have shown significant promise across various medical applications, with ophthalmology being a notable area of focus. Many ophthalmic tasks have shown substantial improvement through the integration of LLMs. However, before these models can be widely adopted in clinical practice, evaluating their capabilities and identifying their limitations is crucial. To address this research gap and support the real-world application of LLMs, we introduce the OphthBench, a specialized benchmark designed to assess LLM performance within the context of Chinese ophthalmic practices. This benchmark systematically divides a typical ophthalmic clinical workflow into five key scenarios: Education, Triage, Diagnosis, Treatment, and Prognosis. For each scenario, we developed multiple tasks featuring diverse question types, resulting in a comprehensive benchmark comprising 9 tasks and 591 questions. This comprehensive framework allows for a thorough assessment of LLMs' capabilities and provides insights into their practical application in Chinese ophthalmology. Using this benchmark, we conducted extensive experiments and analyzed the results from 39 popular LLMs. Our evaluation highlights the current gap between LLM development and its practical utility in clinical settings, providing a clear direction for future advancements. By bridging this gap, we aim to unlock the potential of LLMs and advance their development in ophthalmology.

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.

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.

Danli Shi, Xiaolan Chen, Bingjie Yan et al.

Artificial intelligence has shown promise in medical imaging, yet most existing systems lack flexibility, interpretability, and adaptability - challenges especially pronounced in ophthalmology, where diverse imaging modalities are essential. We present EyeAgent, the first agentic AI framework for comprehensive and interpretable clinical decision support in ophthalmology. Using a large language model (DeepSeek-V3) as its central reasoning engine, EyeAgent interprets user queries and dynamically orchestrates 53 validated ophthalmic tools across 23 imaging modalities for diverse tasks including classification, segmentation, detection, image/report generation, and quantitative analysis. Stepwise ablation analysis demonstrated a progressive improvement in diagnostic accuracy, rising from a baseline of 69.71% (using only 5 general tools) to 80.79% when the full suite of 53 specialized tools was integrated. In an expert rating study on 200 real-world clinical cases, EyeAgent achieved 93.7% tool selection accuracy and received expert ratings of more than 88% across accuracy, completeness, safety, reasoning, and interpretability. In human-AI collaboration, EyeAgent matched or exceeded the performance of senior ophthalmologists and, when used as an assistant, improved overall diagnostic accuracy by 18.51% and report quality scores by 19%, with the greatest benefit observed among junior ophthalmologists. These findings establish EyeAgent as a scalable and trustworthy AI framework for ophthalmology and provide a blueprint for modular, multimodal, and clinically aligned next-generation AI systems.

Mertcan Sevgi, Fares Antaki, Abdullah Zafar Khan et al.

Vision impairment and blindness are a major global health challenge where gaps in the ophthalmology workforce limit access to specialist care. We evaluate AMIE, a medically fine-tuned conversational system based on Gemini with integrated web search and self-critique reasoning, using real-world clinical vignettes that reflect scenarios a general ophthalmologist would be expected to manage. We conducted two complementary evaluations: (1) a human-AI interactive diagnostic reasoning study in which ophthalmologists recorded initial differentials and plans, then reviewed AMIE's structured output and revised their answers; and (2) a masked preference and quality study comparing AMIE's narrative outputs with case author reference answers using a predefined rubric. AMIE showed standalone diagnostic performance comparable to clinicians at baseline. Crucially, after reviewing AMIE's responses, ophthalmologists tended to rank the correct diagnosis higher, reached greater agreement with one another, and enriched their investigation and management plans. Improvements were observed even when AMIE's top choice differed from or underperformed the clinician baseline, consistent with a complementary effect in which structured reasoning support helps clinicians re-rank rather than simply accept the model output. Preferences varied by clinical grade, suggesting opportunities to personalise responses by experience. Without ophthalmology-specific fine-tuning, AMIE matched clinician baseline and augmented clinical reasoning at the point of need, motivating multi-axis evaluation, domain adaptation, and prospective multimodal studies in real-world settings.

Pusheng Xu, Yue Wu, Kai Jin et al.

Purpose: To evaluate the accuracy and reasoning ability of DeepSeek-R1 and three other recently released large language models (LLMs) in bilingual complex ophthalmology cases. Methods: A total of 130 multiple-choice questions (MCQs) related to diagnosis (n = 39) and management (n = 91) were collected from the Chinese ophthalmology senior professional title examination and categorized into six topics. These MCQs were translated into English using DeepSeek-R1. The responses of DeepSeek-R1, Gemini 2.0 Pro, OpenAI o1 and o3-mini were generated under default configurations between February 15 and February 20, 2025. Accuracy was calculated as the proportion of correctly answered questions, with omissions and extra answers considered incorrect. Reasoning ability was evaluated through analyzing reasoning logic and the causes of reasoning error. Results: DeepSeek-R1 demonstrated the highest overall accuracy, achieving 0.862 in Chinese MCQs and 0.808 in English MCQs. Gemini 2.0 Pro, OpenAI o1, and OpenAI o3-mini attained accuracies of 0.715, 0.685, and 0.692 in Chinese MCQs (all P<0.001 compared with DeepSeek-R1), and 0.746 (P=0.115), 0.723 (P=0.027), and 0.577 (P<0.001) in English MCQs, respectively. DeepSeek-R1 achieved the highest accuracy across five topics in both Chinese and English MCQs. It also excelled in management questions conducted in Chinese (all P<0.05). Reasoning ability analysis showed that the four LLMs shared similar reasoning logic. Ignoring key positive history, ignoring key positive signs, misinterpretation medical data, and too aggressive were the most common causes of reasoning errors. Conclusion: DeepSeek-R1 demonstrated superior performance in bilingual complex ophthalmology reasoning tasks than three other state-of-the-art LLMs. While its clinical applicability remains challenging, it shows promise for supporting diagnosis and clinical decision-making.

Ching-Yun Wang, Shiuan-Chen Lin, Kao-Jung Chang et al.

Acute myeloid leukemia (AML) establishes an immunosuppressive microenvironment that favors leukemic proliferation. The immune-suppressive cytokines altered antigen processing, and presentation collectively assist AML cells in escaping cytotoxic T-cell surveillance. These CD8+ T cell dysfunction features are emerging therapeutic targets in relapsed/refractory AML patients. Besides, CD8+ T cell exhaustion is a hotspot in recent clinical oncology studies, but its pathophysiology has yet to be elucidated in AML. In this review, we summarize high-quality original studies encompassing the phenotypic and genomic characteristics of T cell exhaustion events in the leukemia progression, emphasize the surface immuno-peptidome that dynamically tunes the fate of T cells to function or dysfunction states, and revisit the biochemical and biophysical properties of type 1 MHC antigen processing mechanism (APM) that pivots in the phenomenon of leukemia antigen dampening.

Tanner J. Ferguson, Brooke Messer, Nicholas Risbrudt et al.

Abstract Introduction The purpose of this study is to evaluate the use of a varenicline solution nasal spray (VNS) for reducing the signs and symptoms of dry eye following laser in situ keratomileusis (LASIK). Methods Subjects electing to undergo LASIK were randomized to VNS (study group) or placebo/vehicle (control group) and initiated treatment with the nasal spray twice daily 28 days prior to surgery with continued treatment for 84 days following LASIK. After initiation of treatment, subjects were seen on the day of surgery and postoperatively on Days 1, 7, 28, 84 (3 months) and 168 (6 months). The primary outcome measure was the mean change in NEI-VFQ-25, a 25-item dry eye questionnaire, from baseline to 3 months. The second primary outcome measure was the mean change in corneal fluorescein staining. Secondary outcome measures included evaluation of tear break-up time, Schirmer testing, tear osmolarity and eye dryness score (EDS). Results Twenty subjects were enrolled in each group and successfully underwent LASIK. Both groups demonstrated an improvement in the National Eye Institute Visual Function Questionnaire (NEI-VFQ) at 3 months. The study group demonstrated improved corneal staining scores at months 1 and 3. Similarly, the study group demonstrated improvement in tear osmolarity scores versus the placebo group at the same time points. Although the study group was numerically greater than placebo for each time point for both corneal staining and tear osmolarity, the differences were not statistically significant for any primary or secondary outcome measures. Conclusion VNS is a dry eye treatment option for patients following LASIK and may have potential benefit for patients hoping to avoid additional topical medications. The results were not statistically significant compared to placebo in this trial, and further investigation of the use of VNS following LASIK in a larger trial would be beneficial.

Zhenyue Qin, Yu Yin, Dylan Campbell et al.

The prevalence of vision-threatening eye diseases is a significant global burden, with many cases remaining undiagnosed or diagnosed too late for effective treatment. Large vision-language models (LVLMs) have the potential to assist in understanding anatomical information, diagnosing eye diseases, and drafting interpretations and follow-up plans, thereby reducing the burden on clinicians and improving access to eye care. However, limited benchmarks are available to assess LVLMs' performance in ophthalmology-specific applications. In this study, we introduce LMOD, a large-scale multimodal ophthalmology benchmark consisting of 21,993 instances across (1) five ophthalmic imaging modalities: optical coherence tomography, color fundus photographs, scanning laser ophthalmoscopy, lens photographs, and surgical scenes; (2) free-text, demographic, and disease biomarker information; and (3) primary ophthalmology-specific applications such as anatomical information understanding, disease diagnosis, and subgroup analysis. In addition, we benchmarked 13 state-of-the-art LVLM representatives from closed-source, open-source, and medical domains. The results demonstrate a significant performance drop for LVLMs in ophthalmology compared to other domains. Systematic error analysis further identified six major failure modes: misclassification, failure to abstain, inconsistent reasoning, hallucination, assertions without justification, and lack of domain-specific knowledge. In contrast, supervised neural networks specifically trained on these tasks as baselines demonstrated high accuracy. These findings underscore the pressing need for benchmarks in the development and validation of ophthalmology-specific LVLMs.

Zihan Li, Diping Song, Zefeng Yang et al.

The need for improved diagnostic methods in ophthalmology is acute, especially in the underdeveloped regions with limited access to specialists and advanced equipment. Therefore, we introduce VisionUnite, a novel vision-language foundation model for ophthalmology enhanced with clinical knowledge. VisionUnite has been pretrained on an extensive dataset comprising 1.24 million image-text pairs, and further refined using our proposed MMFundus dataset, which includes 296,379 high-quality fundus image-text pairs and 889,137 simulated doctor-patient dialogue instances. Our experiments indicate that VisionUnite outperforms existing generative foundation models such as GPT-4V and Gemini Pro. It also demonstrates diagnostic capabilities comparable to junior ophthalmologists. VisionUnite performs well in various clinical scenarios including open-ended multi-disease diagnosis, clinical explanation, and patient interaction, making it a highly versatile tool for initial ophthalmic disease screening. VisionUnite can also serve as an educational aid for junior ophthalmologists, accelerating their acquisition of knowledge regarding both common and underrepresented ophthalmic conditions. VisionUnite represents a significant advancement in ophthalmology, with broad implications for diagnostics, medical education, and understanding of disease mechanisms. The source code is at https://github.com/HUANGLIZI/VisionUnite.

Liping Chen, Ling Sun, Caihong Xue et al.

Abstract Background An increasing prevalence of mental disorders (MDs) has been reported among children and adolescents. However, only few studies have conducted ocular examinations, including those on refractive status, in these groups of patients. Thus, the purpose of this study was to evaluate the refractive status and ocular findings in children and adolescents with MDs compared with matched controls with similar socioeconomic backgrounds. Methods A total of 178 participants with MDs and 200 controls were recruited between April 2021 and May 2022. All the children and adolescents underwent cycloplegic or noncycloplegic autorefraction and retinoscopy, slit-lamp biomicroscopy, and dilated fundus examinations. Ocular alignment was assessed using Hirschberg, Krimsky, or prism cover tests. The prevalence of refractive errors and ocular findings was the main outcome. Results Twenty-seven percent of patients with MDs and 8% of controls had ocular findings, the most common of which were conjunctivitis, keratitis, and trichiasis. For refractive status, 70% (124/178) of patients with MDs had myopia ≤-1.00 DS, and 2% (4/178) had hyperopia ≥+2.00 DS. In the control group, 70% (140/200) of patients had myopia ≤-1.00 DS, and 1% (2/200) had hyperopia ≥+2.00 DS. No differences were observed between the MD and control groups. However, the patients in the MD group (14.25±2.69 years) were significantly more susceptible to strabismus (P<0.05) and amblyopia (P<0.01) than those in the control group (13.65±3.04 years). There was a substantial difference between the two groups in the time spent on screen-based devices (P<0.001). Furthermore, mental retardation (OR=3.286, P<0.01), emotional disorders (OR=2.003, P<0.01), and adjustment disorders (OR=2.629, P<0.01) were associated with an increased risk of amblyopia. Depression (OR =1.362, P<0.01) and emotional disorders (OR=2.205, P<0.01) were associated with a higher prevalence of strabismus. Conclusion Ophthalmological examinations should be performed in children and adolescents with MDs because MDs are associated with a high prevalence of refractive errors and ocular diseases. Detection and intervention of ocular and refractive findings in children and adolescents with MDs are necessary and effective in alleviating the economic burden in healthcare and improving individuals' quality of life

Weihao Gao, Zhuo Deng, Zhiyuan Niu et al.

Large multimodal language models (LMMs) have achieved significant success in general domains. However, due to the significant differences between medical images and text and general web content, the performance of LMMs in medical scenarios is limited. In ophthalmology, clinical diagnosis relies on multiple modalities of medical images, but unfortunately, multimodal ophthalmic large language models have not been explored to date. In this paper, we study and construct an ophthalmic large multimodal model. Firstly, we use fundus images as an entry point to build a disease assessment and diagnosis pipeline to achieve common ophthalmic disease diagnosis and lesion segmentation. Then, we establish a new ophthalmic multimodal instruction-following and dialogue fine-tuning dataset based on disease-related knowledge data and publicly available real-world medical dialogue. We introduce visual ability into the large language model to complete the ophthalmic large language and vision assistant (OphGLM). Our experimental results demonstrate that the OphGLM model performs exceptionally well, and it has the potential to revolutionize clinical applications in ophthalmology. The dataset, code, and models will be made publicly available at https://github.com/ML-AILab/OphGLM.

Zhongxi Qiu, Yan Hu, Heng Li et al.

Segmentation is vital for ophthalmology image analysis. But its various modal images hinder most of the existing segmentation algorithms applications, as they rely on training based on a large number of labels or hold weak generalization ability. Based on Segment Anything (SAM), we propose a simple but effective learnable prompt layer suitable for multiple target segmentation in ophthalmology multi-modal images, named Learnable Ophthalmology Segment Anything (SAM). The learnable prompt layer learns medical prior knowledge from each transformer layer. During training, we only train the prompt layer and task head based on a one-shot mechanism. We demonstrate the effectiveness of our thought based on four medical segmentation tasks based on nine publicly available datasets. Moreover, we only provide a new improvement thought for applying the existing fundamental CV models in the medical field. Our codes are available at \href{https://github.com/Qsingle/LearnablePromptSAM}{website}.

Jason Holmes, Shuyuan Ye, Yiwei Li et al.

Purpose: The performance of three different large language models (LLMS) (GPT-3.5, GPT-4, and PaLM2) in answering ophthalmology professional questions was evaluated and compared with that of three different professional populations (medical undergraduates, medical masters, and attending physicians). Methods: A 100-item ophthalmology single-choice test was administered to three different LLMs (GPT-3.5, GPT-4, and PaLM2) and three different professional levels (medical undergraduates, medical masters, and attending physicians), respectively. The performance of LLM was comprehensively evaluated and compared with the human group in terms of average score, stability, and confidence. Results: Each LLM outperformed undergraduates in general, with GPT-3.5 and PaLM2 being slightly below the master's level, while GPT-4 showed a level comparable to that of attending physicians. In addition, GPT-4 showed significantly higher answer stability and confidence than GPT-3.5 and PaLM2. Conclusion: Our study shows that LLM represented by GPT-4 performs better in the field of ophthalmology. With further improvements, LLM will bring unexpected benefits in medical education and clinical decision making in the near future.

Surabhi Datta, Tasneem Kaochar, Hio Cheng Lam et al.

We introduce an annotated corpus of 600 ophthalmology notes labeled with detailed spatial and contextual information of ophthalmic entities. We extend our previously proposed frame semantics-based spatial representation schema, Rad-SpatialNet, to represent spatial language in ophthalmology text, resulting in the Eye-SpatialNet schema. The spatially-grounded entities are findings, procedures, and drugs. To accurately capture all spatial details, we add some domain-specific elements in Eye-SpatialNet. The annotated corpus contains 1715 spatial triggers, 7308 findings, 2424 anatomies, and 9914 descriptors. To automatically extract the spatial information, we employ a two-turn question answering approach based on the transformer language model BERT. The results are promising, with F1 scores of 89.31, 74.86, and 88.47 for spatial triggers, Figure, and Ground frame elements, respectively. This is the first work to represent and extract a wide variety of clinical information in ophthalmology. Extracting detailed information can benefit ophthalmology applications and research targeted toward disease progression and screening.