Ophthalmology science最新文献

{"title":"Longitudinal Changes in Corneal Thickness over 8 Years: Findings from the National Institute for Longevity Sciences–Longitudinal Study of Aging Population-Based Cohort Study in Japan","authors":"Hideki Fukuoka MD, PhD ,&nbsp;Chikako Tange PhD ,&nbsp;Fujiko Ando MD, PhD ,&nbsp;Hiroshi Shimokata MD, PhD ,&nbsp;Yukiko Nishita PhD ,&nbsp;Rei Otsuka PhD","doi":"10.1016/j.xops.2025.100860","DOIUrl":"10.1016/j.xops.2025.100860","url":null,"abstract":"<div><h3>Purpose</h3><div>To evaluate age-related changes in central corneal thickness (CCT) and investigate its relationship with other ocular parameters in community-dwelling Japanese adults through an 8-year longitudinal analysis.</div></div><div><h3>Design</h3><div>A population-based, prospective longitudinal cohort study with baseline measurements from 1997 to 2000 and follow-up from 2006 to 2008.</div></div><div><h3>Subjects</h3><div>A total of 631 community-dwelling Japanese adults aged 40 to 79 years (mean age: 55.7 ± 9.7 years) were enrolled from the National Institute for Longevity Sciences–Longitudinal Study of Aging. We excluded participants with corneal pathologies, contact lens use, glaucoma medication, or missing endothelial cell density measurements.</div></div><div><h3>Methods</h3><div>Central corneal thickness was measured using calibrated specular microscopy (SP-2000; Topcon Corporation) at 2 time points approximately 8 years apart. Secondary measurements included corneal endothelial cell density, coefficient of variation in cell size, and corneal curvature. Mixed-effects models analyzed CCT changes, adjusting for sex, season, corneal endothelial cell density, and systemic health factors.</div></div><div><h3>Main Outcome Measures</h3><div>Age-related changes in CCT, annual rate of CCT change across different age decades, and correlations between CCT changes and ocular/systemic parameters.</div></div><div><h3>Results</h3><div>At baseline, adjusted CCT measurements were 520.2 ± 2.1 (standard error [SE]) μm, 514.1 ± 2.2 μm, 518.0 ± 2.5 μm, and 514.7 ± 3.7 μm for participants in their 40s, 50s, 60s, and 70s, respectively. Longitudinal analysis revealed a significant increase in CCT over time across all age groups (β = 0.7; SE = 0.1; <em>P</em> &lt; 0.001). The annual CCT increase showed age-dependent slowing: 0.68 ± 0.08 μm for 40s, 0.62 ± 0.08 μm for 50s, 0.46 ± 0.09 μm for 60s, and 0.20 ± 0.14 μm for 70s with a statistically significant difference between 40s and 70s groups (β = −0.5; SE = 0.2' <em>P</em> = 0.003).</div></div><div><h3>Conclusions</h3><div>This longitudinal analysis demonstrates that CCT increases over time in all age groups, with the rate of increase significantly slowing in older age groups. These findings contrast with previous cross-sectional studies suggesting CCT decreases with age, emphasizing the importance of longitudinal observations. These results have important implications for glaucoma diagnosis and refractive surgery safety evaluations in aging populations.</div></div><div><h3>Financial Disclosure(s)</h3><div>The author(s) have no proprietary or commercial interest in any materials discussed in this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100860"},"PeriodicalIF":3.2,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144678997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Applications of Computer Vision for Infectious Keratitis: A Systematic Review","authors":"Jad F. Assaf MD ,&nbsp;Abhimanyu S. Ahuja MD ,&nbsp;Vishnu Kannan ,&nbsp;Hady Yazbeck MD ,&nbsp;Jenna Krivit MD ,&nbsp;Travis K. Redd MD, MPH","doi":"10.1016/j.xops.2025.100861","DOIUrl":"10.1016/j.xops.2025.100861","url":null,"abstract":"<div><h3>Clinical Relevance</h3><div>Corneal ulcers cause preventable blindness in &gt;2 million individuals annually, primarily affecting low- and middle-income countries. Prompt and accurate pathogen identification is essential for targeted antimicrobial treatment, yet current diagnostic methods are costly and slow and require specialized expertise, limiting accessibility.</div></div><div><h3>Methods</h3><div>We systematically reviewed literature published from 2017 to 2024, identifying 37 studies that developed or validated artificial intelligence (AI) models for pathogen detection and related classification tasks in infectious keratitis. The studies were analyzed for model types, input modalities, datasets, ground truth determination methods, and validation practices.</div></div><div><h3>Results</h3><div>Artificial intelligence models demonstrated promising accuracy in pathogen detection using image interpretation techniques. Common limitations included limited generalizability, lack of diverse datasets, absence of multilabeled classification methods, and variability in ground truth standards. Most studies relied on single-center retrospective datasets, limiting applicability in routine clinical practice.</div></div><div><h3>Conclusions</h3><div>Artificial intelligence shows significant potential to improve pathogen detection in infectious keratitis, enhancing both diagnostic accuracy and accessibility globally. Future research should address identified limitations by increasing dataset diversity, adopting multilabel classification, implementing prospective and multicenter validations, and standardizing ground truth definitions.</div></div><div><h3>Financial Disclosure(s)</h3><div>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100861"},"PeriodicalIF":4.6,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144722800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and Evaluation of a Computable Phenotype for Normal Tension Glaucoma","authors":"Fountane Chan MD ,&nbsp;Wei-Chun Lin MD, PhD ,&nbsp;Alan Tang ,&nbsp;Benjamin Y. Xu MD, PhD ,&nbsp;Sophia Y. Wang MD, MS ,&nbsp;Michael V. Boland MD, PhD ,&nbsp;Catherine Q. Sun MD ,&nbsp;Sally Baxter MD, MSc ,&nbsp;Brian Stagg MD, MS ,&nbsp;Michelle Hribar PhD ,&nbsp;Aiyin Chen MD","doi":"10.1016/j.xops.2025.100858","DOIUrl":"10.1016/j.xops.2025.100858","url":null,"abstract":"<div><h3>Purpose</h3><div>To develop a computable phenotype for normal tension glaucoma (NTG) to enhance disease identification from electronic health records (EHRs).</div></div><div><h3>Design</h3><div>Retrospective cohort study.</div></div><div><h3>Subjects</h3><div>Deidentified EHR data from an academic medical center identified 1851 patients aged ≥40 years, with glaucoma and available clinical notes.</div></div><div><h3>Methods</h3><div>Of these 1851 patients, 200 were randomly selected for a chart review to receive gold standard diagnoses. Four rule-based NTG computable phenotypes were developed and tested. Phenotype 1 relied on NTG International Classification of Diseases (ICD)-9 and ICD-10 codes. Phenotype 2 incorporated structured intraocular pressure (IOP) data and medication lists. Phenotype 3 used only structured IOP data. Phenotype 4 combined structured IOP and medication data natural language processing (NLP) to extract IOP values and NTG mentions from chart notes. Internal and external validation were performed.</div></div><div><h3>Main Outcome Measures</h3><div>F1 score, sensitivities, specificities, positive predictive value (PPV), negative predictive value (NPV), and accuracy.</div></div><div><h3>Results</h3><div>Chart review identified NTG in 30% of patients, and only 7% had NTG ICD codes. Phenotype 1 had an F1 of 36.8%, sensitivity 24.1%, specificity 97%, PPV 77.8%, NPV 74.9%, and accuracy 75.1%. Compared with ICD codes, phenotypes 2 and 3 had F1 of 66.7% and 69.8%, sensitivity 77.6% and 89.7%, specificity 76.3% and 71.1%, PPV 58.4% and 57.1%, NPV 88.8% and 94.1%, and accuracy of 76.7% and 76.7%, respectively. Incorporating NLP, phenotype 4 had the best performance with an F1 of 77.4%, sensitivity 82.8%, specificity 86.7%, PPV 72.7%, NPV 92.1%, and accuracy 85.5%. Phenotypes 2 to 4 increase NTG case detection fourfold compared with phenotype 1.</div></div><div><h3>Conclusions</h3><div>Normal tension glaucoma phenotypes using NLP achieved the best overall performance, and those incorporating structured data perform better than ICD codes alone. The NTG ICD code-based phenotype is highly specific but lacks sensitivity. Insights from this study may inform the development of computable phenotypes for other disease subtypes within broader disease categories.</div></div><div><h3>Financial Disclosure(s)</h3><div>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100858"},"PeriodicalIF":3.2,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144671087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predicting Progression to Vision-Threatening Complications in Diabetic Retinopathy","authors":"C. Quentin Davis PhD ,&nbsp;Nadia K. Waheed MD ,&nbsp;Mitchell Brigell PhD","doi":"10.1016/j.xops.2025.100859","DOIUrl":"10.1016/j.xops.2025.100859","url":null,"abstract":"<div><h3>Objective</h3><div>To characterize the performance of 56 parameters from electroretinography (ERG)/pupillometry, color fundus photography (FP), OCT angiography (OCTA), and ultra-widefield fluorescein angiography (UWF-FA) for predicting which subjects with nonproliferative diabetic retinopathy (NPDR) will progress to vision-threating complications (VTCs) within 48 weeks.</div></div><div><h3>Design</h3><div>A longitudinal prospective study from 44 trial sites in the United States.</div></div><div><h3>Participants</h3><div>Subjects had moderate-to-severe NPDR and no center-involved diabetic macular edema. Among the 162 subjects, the mean age was 57 years and 58% were male.</div></div><div><h3>Intervention</h3><div>Although this study tested an experimental drug, there was no indication of a treatment effect. Results are analyzed over all subjects regardless of study treatment.</div></div><div><h3>Main Outcome Measures</h3><div>Specialized reading centers measured 56 parameters from 4 testing modalities (ERG/pupillometry, FP, OCTA, and UWF-FA) to evaluate diabetic retinopathy (DR) status. Kaplan–Meier analysis and a Cox proportional hazards model were applied to each parameter to identify significant predictors of progression to VTC, defined as progression to proliferative DR, diabetic macular edema, or treatment thereof.</div></div><div><h3>Results</h3><div>Of the 56 parameters, the strongest predictor of progression in the following 48 weeks was the RETeval DR score, which combines ERG and pupil response. A DR score ≥26.9 had a relative risk (RR) of 5.6 (<em>P</em> &lt; 0.0001). The most predictive parameter from the other modalities were UWF-FA's total ischemia index ≥0.125 with an RR of 5.3 (<em>P</em> &lt; 0.0001), OCTA's foveal avascular zone area ≥0.295 mm² with an RR of 3.6 (<em>P</em> &lt; 0.05), and FP's diabetic retinopathy severity scale ≥47 (moderate NPDR) with an RR of 2.1 (<em>P</em> &lt; 0.05).</div></div><div><h3>Conclusions</h3><div>Both functional (ERG, pupil response) and structural (FP, OCTA, UWF-FA) testing can predict progression to VTC from DR, with the DR score having the best predictive capability. These results suggest it is possible to improve the DR staging system which in turn may enable better allocation of health care resources.</div></div><div><h3>Financial Disclosure(s)</h3><div>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100859"},"PeriodicalIF":4.6,"publicationDate":"2025-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144722869","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Long-Shaft Vitrectomy Probe for Vitreoretinal Diseases in Highly Myopic Eyes: A Randomized Controlled Trial","authors":"Yun Hsia MD, MS ,&nbsp;Cheng-Yung Lee MD ,&nbsp;Mei-Chi Tsui MD ,&nbsp;Shih-Wen Wang MD ,&nbsp;Chien-Jung Huang MD ,&nbsp;I-Hsin Ma MD, MS ,&nbsp;Kuo-Chi Hung MD ,&nbsp;Muh-Shy Chen MD, PhD ,&nbsp;Zih-Wei Yang MD ,&nbsp;Bo-Da Huang MD ,&nbsp;Ting-Chieh Ko MD ,&nbsp;Tzyy-Chang Ho MD","doi":"10.1016/j.xops.2025.100824","DOIUrl":"10.1016/j.xops.2025.100824","url":null,"abstract":"<div><h3>Objective</h3><div>To evaluate the safety and efficacy of a long-shaft vitrectomy probe in highly myopic eyes undergoing pars plana vitrectomy.</div></div><div><h3>Design</h3><div>A randomized controlled trial.</div></div><div><h3>Subjects</h3><div>Highly myopic eyes (axial length [AL] &gt;26 mm) with epiretinal membrane (ERM), myopic tractional maculopathy, and retinal detachment.</div></div><div><h3>Methods</h3><div>The enrolled eyes were randomly assigned to a study group (UltraVit 25 Ga, 31.75 mm, Alcon) and a control group (UltraVit 25 Ga+, 27 mm, Alcon). Stratified randomization was performed to balance the proportion of eyes with AL &gt;28 mm between groups.</div></div><div><h3>Main Outcome Measures</h3><div>Trocar removal rate and core vitrectomy time were assessed as primary outcomes, and instrument bending as a secondary outcome. Anatomical and visual outcomes and complications were documented for 6 months. Subgroup analysis was performed to compare the eyes with AL &gt;28 mm to those without.</div></div><div><h3>Results</h3><div>We included 86 patients with a mean age of 60.7 ± 9.6 years and an AL of 29.15 ± 2.14 mm. Two groups had comparable core vitrectomy times (−0.5 minutes, <em>P</em> = 0.172). The study group had a lower trocar removal rate (5% vs. 67%, <em>P</em> &lt; 0.001) but a higher instrument bending rate (36% vs. 14%, <em>P</em> = 0.036), particularly in eyes with AL &gt;28 mm. In eyes with AL &gt;28 mm, the standard vitrectomy probe faced a significantly greater difficulty in cortical vitreous removal, internal limiting membrane (ILM) trimming, or posterior vitreous detachment induction compared with the long-shaft vitrectomy (<em>P</em> &lt; 0.001). At 6 months, significant visual improvement (logarithm of the minimum angle of resolution) and anatomical success were achieved (study: −0.22, 95%; control: −0.24, 88%). Eyes with ERM in the study group, not the controls, had significant visual improvement (−0.21, <em>P</em> = 0.02 vs. −0.09, <em>P</em> = 0.34).</div></div><div><h3>Conclusions</h3><div>The long-shaft vitrectomy probe is safe and efficient, especially in highly myopic eyes with AL &gt;28 mm, despite a higher instrument bending rate. It provides improved access to the posterior pole, allowing for delicate removal of adherent cortical vitreous and trimming of ILM flaps. Addressing increased instrument bending due to the sleeveless design is important for future design.</div></div><div><h3>Financial Disclosure(s)</h3><div>The author(s) have no proprietary or commercial interest in any materials discussed in this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 5","pages":"Article 100824"},"PeriodicalIF":3.2,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144338967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integration of Immunometabolic Composite Indices and Machine Learning for Diabetic Retinopathy Risk Stratification: Insights from NHANES 2011 – 2020","authors":"Cui Xuehao MD, PhD ,&nbsp;Wen Dejia MD ,&nbsp;Li Xiaorong PhD","doi":"10.1016/j.xops.2025.100854","DOIUrl":"10.1016/j.xops.2025.100854","url":null,"abstract":"<div><h3>Objective</h3><div>This study aimed to investigate the association between immunometabolic composite indices and diabetic retinopathy (DR) and to develop predictive models using machine learning (ML) techniques to improve early detection and risk stratification for DR.</div></div><div><h3>Design</h3><div>A cross-sectional study.</div></div><div><h3>Subjects and Controls</h3><div>Data from the National Health and Nutrition Examination Survey 2011–2020 were analyzed, involving 8249 participants categorized into healthy controls (n = 6830), diabetes without retinopathy (n = 918), and DR (n = 501).</div></div><div><h3>Methods</h3><div>Immunometabolic indices reflecting insulin resistance, inflammation, and lipid metabolism were evaluated. Multivariate logistic regression models assessed associations with DR, and Bayesian kernel machine regression analyzed nonlinear interactions. Eight ML models, including ensemble methods, were developed to predict DR risk, with feature importance determined by SHapley Additive exPlanations.</div></div><div><h3>Main Outcome Measures</h3><div>The primary outcome was DR status, classified according to the ETDRS criteria from fundus photography.</div></div><div><h3>Results</h3><div>Key immunometabolic indices, notably Frailty Index (FRAILTY) and fasting serum insulin (FSI), were significantly associated with increased DR risk, whereas the metabolic score for insulin resistance (METS) showed a protective effect. Bayesian kernel machine regression highlighted complex interactions among indices. Machine learning models achieved high predictive accuracy, particularly XGBoost and LightGBM (area under the curve &gt; 0.9). SHapley Additive exPlanations analyses identified FRAILTY, FSI, and METS as the most influential predictors.</div></div><div><h3>Conclusions</h3><div>Immunometabolic dysregulation significantly contributes to DR progression beyond traditional risk factors such as hyperglycemia alone. Incorporating immunometabolic indices into predictive models substantially enhances DR risk stratification, facilitating personalized screening and intervention strategies. Machine learning approaches effectively identify high-risk individuals, underscoring their utility in clinical practice for early DR detection and targeted preventive care.</div></div><div><h3>Financial Disclosure(s)</h3><div>The author(s) have no proprietary or commercial interest in any materials discussed in this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100854"},"PeriodicalIF":4.6,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144725107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Retinal Blood Flow Decreases after Treatment with Bevacizumab for Retinopathy of Prematurity","authors":"Joyce Wang BS ,&nbsp;Shaiza Mansoor BS ,&nbsp;Jeong-Yoon Wu BS ,&nbsp;Christina Kilby BS ,&nbsp;He Forbes MS ,&nbsp;Ria Kapoor BS ,&nbsp;Sarah Ward ,&nbsp;Jason Zhou BS ,&nbsp;Kristin Williams RN ,&nbsp;Moran Roni Levin MD ,&nbsp;Sripriya Sundararajan MD ,&nbsp;Larry Magder PhD ,&nbsp;Avigyan Sinha PhD ,&nbsp;Abhishek Rege PhD ,&nbsp;Janet L. Alexander MD, MS","doi":"10.1016/j.xops.2025.100857","DOIUrl":"10.1016/j.xops.2025.100857","url":null,"abstract":"<div><h3>Purpose</h3><div>To compare total retinal blood flow (TRBF) rates before and after retinopathy of prematurity (ROP) treatment with intravitreal bevacizumab using laser speckle contrast imaging (LSCI).</div></div><div><h3>Design</h3><div>A prospective cohort study.</div></div><div><h3>Participants</h3><div>Twenty-five eyes from 14 premature infants in the neonatal intensive care unit receiving intravitreal bevacizumab for treatment-requiring ROP.</div></div><div><h3>Methods</h3><div>Total retinal blood flow was measured using LSCI longitudinally before and after bevacizumab treatment. Subject characteristics and clinical ROP features, including the need for ROP retreatment, were included in regression analysis using generalized estimating equations to account for 2 eyes per subject and longitudinal measures over time.</div></div><div><h3>Main Outcome Measures</h3><div>The main outcome measure was TRBF, which includes components of peak, mean, and dip over the cardiac cycle.</div></div><div><h3>Results</h3><div>Before ROP treatment, subjects had a peak TRBF of 11.1 ± 2.9 a.u. compared to 8.6 ± 1.8 a.u. after treatment (mean difference = 2.5 a.u., <em>P</em> &lt; 0.0001). Among eyes that required ROP retreatment earlier (&lt;10 weeks) after initial treatment, the posttreatment peak TRBF was 9.0 ± 1.5 a.u., compared to 7.3 ± 2.2 a.u. for eyes that did not require retreatment in the first 10 weeks after initial bevacizumab injection (mean difference = 1.7 a.u., <em>P</em> = 0.01). Peak TRBF decreased over time after bevacizumab treatment (β = −0.1 a.u./week, <em>P</em> = 0.004).</div></div><div><h3>Conclusions</h3><div>We observed lower TRBF after treatment with intravitreal bevacizumab.</div></div><div><h3>Financial Disclosure(s)</h3><div>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100857"},"PeriodicalIF":3.2,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144632090","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated Segmentation of Subretinal Fluid from OCT: A Vision Transformer Approach with Cross-Validation","authors":"Julie Midroni HBSc ,&nbsp;Jack Longwell HBSc, MEng(C) ,&nbsp;Nishaant Bhambra MD ,&nbsp;Sueellen Demian MD ,&nbsp;Aurora Pecaku MD ,&nbsp;Isabela Martins Melo MD ,&nbsp;Rajeev H. Muni MD, MSc","doi":"10.1016/j.xops.2025.100852","DOIUrl":"10.1016/j.xops.2025.100852","url":null,"abstract":"<div><h3>Purpose</h3><div>We present an algorithm to segment subretinal fluid (SRF) on individual B-scan slices in patients with rhegmatogenous retinal detachment (RRD). Particular attention is paid to robustness, with a fivefold cross-validation approach and a hold-out test set.</div></div><div><h3>Design</h3><div>Retrospective, cross-sectional study.</div></div><div><h3>Participants</h3><div>A total of 3819 B-scan slices across 98 time points from 45 patients were used in this study.</div></div><div><h3>Methods</h3><div>Subretinal fluid was segmented on all scans. A base SegFormer model, pretrained on 4 massive data sets, was further trained on raw B-scans from the retinal OCT fluid challenge data set of 4532 slices: an open data set of intraretinal fluid, SRF, and pigment epithelium detachment. When adequate performance was reached, transfer learning was used to train the model on our in-house data set, to segment SRF by generating a pixel-wise mask of presence/absence of SRF. A fivefold cross-validation approach was used, with an additional hold-out test set. All folds were first trained and cross-validated and then additionally tested on the hold-out set. Mean (averaged across images) and total (summed across all pixels, irrespective of image) Dice coefficients were calculated for each fold.</div></div><div><h3>Main Outcome Measures</h3><div>Subretinal fluid volume after surgical intervention for RRD.</div></div><div><h3>Results</h3><div>The average total Dice coefficient across the validation folds was 0.92, the average mean Dice coefficient was 0.82, and the median Dice was 0.92. For the test set, the average total Dice coefficient was 0.94, the average mean Dice coefficient was 0.82, and the median Dice was 0.92. The model showed strong interfold consistency on the hold-out set, with a standard deviation of only 0.03.</div></div><div><h3>Conclusions</h3><div>The SegFormer model for SRF segmentation demonstrates a strong ability to segment SRF. This result holds up to cross-validation and hold-out testing, across all folds. The model is available open-source online.</div></div><div><h3>Financial Disclosure(s)</h3><div>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100852"},"PeriodicalIF":4.6,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144722789","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Genetic Risk and OCT-Based Phenotypic Associations in Age-Related Macular Degeneration","authors":"Shlomit Jaskoll ,&nbsp;Yahel Shwartz MSc ,&nbsp;Adi Kramer MSc ,&nbsp;Sarah Elbaz-Hayoun PhD ,&nbsp;Batya Rinsky PhD ,&nbsp;Michelle Grunin PhD ,&nbsp;Liran Tiosano MD ,&nbsp;Jaime Levy MD ,&nbsp;Brice Nguedia Vofo MD, MBA ,&nbsp;Itay Chowers MD","doi":"10.1016/j.xops.2025.100853","DOIUrl":"10.1016/j.xops.2025.100853","url":null,"abstract":"<div><h3>Purpose</h3><div>The risk for developing age-related macular degeneration (AMD) is associated with multiple genetic variants. We aim to evaluate the association of AMD genetic risk variants with specific features of the disease detected by OCT.</div></div><div><h3>Design</h3><div>A retrospective cross-sectional study.</div></div><div><h3>Participants</h3><div>Subjects diagnosed with AMD and healthy controls (&gt;50 years of age) from a single tertiary referral center.</div></div><div><h3>Methods</h3><div>Genotyping of 52 single nucleotide polymorphisms associated with AMD was analyzed in 578 patients. Weighted genetic risk scores (WGRSs) were calculated for variants in genes encoding proteins involved in the complement cascade, lipid metabolism, and other pathways, respectively. A global WGRS was calculated for all 52 variants. OCT images were annotated for the presence of typical drusen, subretinal drusenoid deposits, hyperreflective foci (HRF), complete retinal pigmented epithelium and outer retinal atrophy (cRORA), and macular neovascularization.</div></div><div><h3>Main Outcome Measures</h3><div>Association of WGRS and individual genetic risk variants with specific disease features detected by OCT.</div></div><div><h3>Results</h3><div>A positive correlation between the presence of drusen and the lipid WGRS was detected (<em>r</em> = 0.09, <em>P</em> = 0.02). Logistic regression analysis indicated associations between cRORA and the complement score (odds ratio [OR] = 1.25, 95% confidence interval [CI] 1.05–1.50; <em>P</em> = 0.01), as well as the global score (OR = 1.29, 95% CI 1.13–1.46; <em>P</em> &lt; 0.001). Regression also showed an association of HRF with the age-related maculopathy susceptibility 2/high-temperature requirement A serine peptidase 1 variant (OR = 1.53, 95% CI 1.03–2.27; <em>P</em> = 0.03), the other pathways score (OR = 1.94, 95% CI 1.20–3.12; <em>P</em> = 0.007), and the global score (OR = 1.16, 95% CI 1.00–1.35; <em>P</em> = 0.04).</div></div><div><h3>Conclusions</h3><div>Weighted genetic risk scores based on risk variants for AMD are associated with specific disease features. Tighter association of the global WGRS compared to pathway-specific scores suggests that several pathways are involved in the development of specific disease features such as cRORA, drusen, and HRF.</div></div><div><h3>Financial Disclosure(s)</h3><div>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100853"},"PeriodicalIF":3.2,"publicationDate":"2025-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144632089","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Using Machine Learning to Identify Ophthalmology Subspecialty Care and Advance Workforce Research with the IRIS® Registry (Intelligent Research in Sight)","authors":"Ju Hyun Jeon MS ,&nbsp;Ju-Yeun Lee MD, PhD ,&nbsp;Tobias Elze PhD ,&nbsp;Joan W. Miller MD ,&nbsp;Alice C. Lorch MD, MPH ,&nbsp;Mei-Sing Ong PhD ,&nbsp;Ann Chen Wu MD, MPH ,&nbsp;David G. Hunter MD, PhD ,&nbsp;Isdin Oke MD, MPH","doi":"10.1016/j.xops.2025.100855","DOIUrl":"10.1016/j.xops.2025.100855","url":null,"abstract":"<div><h3>Purpose</h3><div>To develop machine-learning models to identify ophthalmology subspecialists using deidentified patient data from a large database.</div></div><div><h3>Design</h3><div>Cross-sectional.</div></div><div><h3>Participants</h3><div>All ophthalmologists participating in the American Academy of Ophthalmology's IRIS® Registry (Intelligent Research in Sight) from 2013 to 2023 were classified under one of the following general or subspecialty categories: comprehensive, cataract, cornea, glaucoma, retina, oculofacial, pediatric, or neuro-ophthalmology.</div></div><div><h3>Methods</h3><div>We collected the diagnosis, procedure, and prescription codes linked to each ophthalmologist. We performed binary subspecialty classification using random forest models with fivefold cross validation and multispecialty classification using 4 approaches (diagnosis only, procedure only, prescription only, and combined).</div></div><div><h3>Main Outcome Measures</h3><div>Model performance was assessed using area under the receiver operating characteristic curve (AUROC), F1 scores, and Matthews correlation coefficient.</div></div><div><h3>Results</h3><div>The study included 9032 ophthalmologists. Classification accuracy differed by subspecialty (AUROC, retina: 0.981; oculofacial: 0.975; pediatric: 0.972; glaucoma: 0.937; cornea: 0.932; neuro: 0.912; cataract: 0.861; and comprehensive: 0.760). The procedure-only random forest model had better performance (AUROC, 0.903) than the diagnosis-only (0.880) and prescription-only (0.835) model.</div></div><div><h3>Conclusions</h3><div>Machine learning models leveraging the IRIS Registry can provide a near real-time assessment of the landscape of ophthalmic subspecialty care. Identifying subspecialty physicians through practice patterns may provide valuable insights into the future trends of eye care delivery with implications for workforce research and policy interventions.</div></div><div><h3>Financial Disclosure(s)</h3><div>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</div></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 6","pages":"Article 100855"},"PeriodicalIF":3.2,"publicationDate":"2025-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144662359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}