Cleveland Eye Clinic

This cohort study examines the association between cataract surgery and progression of age-related macular degeneration in adults 60 years or older.

To describe swept-source OCT and OCT angiography (OCTA) B-scan findings in inflammatory retinal vascular changes. Observational cross sectional study of prospectively enrolled patients. A total of 87 eyes from 50 patients with inflammatory retinal vascular changes. Swept-source OCTA, which provides both structural OCT and OCTA B-scan data, and ultra-widefield fluorescein angiography were obtained on the same day or within 2 weeks. Structural OCT and OCTA B-scans corresponding to vessel points with inflammatory findings on fluorescein angiography were analyzed for vascular and perivascular structural changes. In eyes with comparable active and inactive phases, B-scan parameters were measured and their changes were compared. Retinal thickness and elevation ratio, outer nuclear layer (ONL) thinning ratio, horizontal and vertical flow signal diameters of retinal veins and arteries, and perivascular changes. The mean age was 42.9 years. Hyporeflective and/or hyperreflective perivascular changes were observed around the flow signal in the active phase. The mean retinal thickness significantly decreased from 318.5 μm in the active phase to 260.3 μm in the inactive phase (P < 0.01) for veins, and 300.5 to 247.0 μm (P < 0.01) for arteries. The mean vertical flow signal diameter for veins decreased from 154.5 to 134.5 μm (P < 0.01) and, for arteries, from 104.8 to 92.0 μm (P = 0.07). The mean horizontal flow signal diameters for veins changed from 145.0 to 130.1 μm (P = 0.07) and, for arteries, from 80.1 to 79.5 μm (P = 0.72). The mean retinal elevation ratio for veins significantly decreased from 1.15 to 1.06 (P < 0.01) and, for arteries, 1.09 to 1.05 (P = 0.08). The mean ONL thinning ratio for veins was 0.74 in the active phase and 0.81 in the inactive phase (P < 0.01), and, for arteries, 0.90 and 0.95 (P = 0.90), respectively. Optical coherence tomography and OCTA B-scan data identified significant changes in retinal thickness, elevation ratio, ONL thinning ratio, and vertical flow signal diameter of veins between the active and inactive phases of inflamed retinal vessels. Additionally, OCT B-scan identified unique perivascular changes, suggesting its potential in identifying inflammatory activity. Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

PurposeCurrent benchmarks evaluating large language models (LLMs) in ophthalmology are narrow and disproportionately prioritize accuracy. We introduce BEnchmarking LLMs for Ophthalmology (BELO), a standardized evaluation benchmark developed through multiple rounds of expert checking by 13 ophthalmologists. BEnchmarking LLMs for Ophthalmology assesses ophthalmology-related knowledge and reasoning quality.SubjectsThis study did not involve human participation.DesignCross-sectional study.MethodsUsing keyword matching and a fine-tuned PubMed Bidirectional Encoder Representations from Transformers model, we curated ophthalmology-specific multiple-choice questions (MCQs) from diverse medical data sets (Basic and Clinical Science Course [BCSC], Multi-Subject Multi-Choice Dataset for Medical domain [MedMCQA], Medical Question Answering [MedQA], Biomedical Semantic Indexing and Question Answering [BioASQ], and PubMed Question Answering [PubMedQA]). The data set 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 3 senior ophthalmologists. To illustrate BELO's utility, we evaluated 8 LLMs (OpenAI o1, o3-mini, GPT-5, GPT-4o, DeepSeek-R1, MedGemma-4B, Llama-3-8B, and Gemini 1.5 Pro).Main Outcome MeasuresThe 8 LLMs were evaluated in terms using accuracy, macro-F1, and 5 text-generation metrics (Recall-Oriented Understudy for Gisting Evaluation, BERTScore, BARTScore, Metric for Evaluation of Translation with Explicit Ordering, and AlignScore). In a further evaluation involving human experts, 2 ophthalmologists qualitatively reviewed 50 randomly selected outputs for accuracy, comprehensiveness, and completeness.ResultsBEnchmarking LLMs for Ophthalmology consists of 900 high-quality, expert-reviewed questions aggregated from 5 sources: BCSC (260), BioASQ (10), MedMCQA (572), MedQA (40), and PubMedQA (18). To demonstrate BELO's utility, we conducted a series of benchmarking exercises. In the quantitative evaluation, GPT-5 achieved the highest accuracy (0.90, 95% confidence interval [CI]: 0.89–0.92) and macro-F1 score (0.91, 95% CI: 0.89–0.93). On the other hand, the models' performance on text-generation metrics varied and were generally suboptimal, with scores ranging from 20.4 to 72.0 (out of 100, excluding the BARTScore metric), indicating room for improvement in clinical reasoning. In expert evaluations, GPT-4o was rated highest for accuracy and readability, while Gemini 1.5 Pro scored highest for completeness. A public leaderboard has been established to promote transparent evaluation and reporting. Importantly, the BELO data set will remain a hold-out, evaluation-only benchmark to ensure fair and reproducible comparisons of future models.ConclusionsBEnchmarking LLMs for Ophthalmology provides a robust clinically relevant benchmark for evaluating both the accuracy and reasoning capabilities of current and emerging LLMs in ophthalmology. Future BELO benchmarking efforts will expand to include vision-based question answering and clinical scenario management tasks.Financial Disclosure(s)The authors have no proprietary or commercial interest in any materials discussed in this article.

The field of visual and physiological optics is undergoing continuous significant advancements, driven by a deeper understanding of the human visual system and the development of cutting-edge optical technologies. This Roadmap, authored by leading experts, delves into critical areas such as corneal biomechanical properties, keratoconus, and advancements in corneal imaging and elastography. It explores the intricate structure-function relationship within the eye lens, offering new perspectives through lens models and ray tracing techniques. The document also covers advancements in retinal imaging, highlighting the current state and future directions, and the role of adaptive optics in evaluating retinal structure and function in both healthy and diseased eyes. Furthermore, it addresses the modeling of ocular surfaces using different mathematical functions and examines the factors affecting peripheral image quality in the human eye, emphasizing the importance of these aspects in visual performance. Additional topics include schematic and functional models of the human eye, the impact of optical and chromatic aberrations, and the design of contact, and intraocular lenses. Finally, the Roadmap addresses the intersection of neurosciences with vision health, presenting a comprehensive overview of current research and future trends aimed at improving visual health and optical performance. Ultimately, this Roadmap aims to serve as a valuable resource for ophthalmologists, optometrists, vision scientists, and engineers dedicated to advancing the field of visual and physiological optics.

To develop consensus-based imaging guidelines for diagnosing and monitoring birdshot chorioretinopathy (BSCR). Consensus-based approach guided by literature and an expert committee using a nominal group technique (NGT). An expert committee of 5 international uveitis specialists reviewed 15 well-documented representative BSCR cases with comprehensive imaging data. Cases with active and inactive disease were included. Imaging, including color fundus photography (CFP), fundus fluorescein angiography (FFA), indocyanine green angiography (ICGA), optical coherence tomography (OCT), fundus autofluorescence (FAF), and OCT angiography (OCTA) were reviewed. Using a structured NGT approach, consensus-based recommendations were developed for specific disease characteristics, biomarkers of activity, and complications. The recommendations were voted upon by members of the full task force. For the diagnosis of BSCR, CFP, FFA, and ICGA were identified as the key imaging modalities. ICGA was identified as a key imaging modality for assessing the presence of choroidal lesions. FFA was deemed crucial for monitoring retinal vascular leakage and assessing the treatment response. OCT, while not essential for diagnosis, was valuable for detecting complications such as cystoid macular edema and retinal thinning. The committee did not reach a consensus on the role of FAF and OCTA for the diagnosis or monitoring of BSCR. The MUV consensus-based imaging guidelines for BSCR expand the Standardization of Uveitis Nomenclature (SUN) classification criteria by reaffirming the critical role of ICGA and providing a standardized guidelines for using other imaging modalities in the diagnosis and monitoring of BSCR. These guidelines are expected to facilitate monitoring of disease activity and complications using multimodal imaging.

Electroretinography (ERG) provides a noninvasive functional measure of multiple cell types of the outer retina. We conducted an ERG-based screen of 530 single-gene knockout mouse strains generated as part of the International Mouse Phenotyping Consortium, representing 2.5% of all protein-coding genes, to identify genetic variants affecting retinal function. We identified 30 strains with significantly altered ERG amplitudes. Two of the genes identified, Cfap418 and Syne2, have been previously reported with outer retinal dysfunction, thereby serving as internal controls that validate our screening protocol. Of the remaining 28 genes newly associated with altered retinal function, the majority lacked a contemporaneous histopathology correlate, highlighting the importance of ERG in early detection of functional abnormalities. A rare homozygous missense variant in FCHSD2, the human orthologue of one of the 28 genes identified, was found in a patient presenting with retinal degeneration that lacked a molecular diagnosis. This report represents a useful resource for future investigations into the molecular mechanisms driving inherited retinal diseases and demonstrates the power of large-scale ERG screening in identifying novel genetic determinants of retinal function.