Ophthalmology science最新文献

{"title":"Retinal Vessel Plexus Differentiation Based on OCT Angiography Using Deep Learning","authors":"Jamie L. Shaffer MS ,&nbsp;Luis De Sisternes PhD ,&nbsp;Anand E. Rajesh BS ,&nbsp;Marian S. Blazes MD ,&nbsp;Yuka Kihara PhD ,&nbsp;Cecilia S. Lee MD, MS ,&nbsp;Warren H. Lewis MS ,&nbsp;Roger A. Goldberg MD ,&nbsp;Niranchana Manivannan PhD ,&nbsp;Aaron Y. Lee MD, MSCI","doi":"10.1016/j.xops.2024.100605","DOIUrl":"10.1016/j.xops.2024.100605","url":null,"abstract":"<div><h3>Purpose</h3><div>Although structural OCT is traditionally used to differentiate the vascular plexus layers in OCT angiography (OCTA), the vascular plexuses do not always obey the retinal laminations. We sought to segment the superficial, deep, and avascular plexuses from OCTA images using deep learning without structural OCT image input or segmentation boundaries.</div></div><div><h3>Design</h3><div>Cross-sectional study.</div></div><div><h3>Subjects</h3><div>The study included 235 OCTA cubes from 33 patients for training and testing of the model.</div></div><div><h3>Methods</h3><div>From each OCTA cube, 3 weakly labeled images representing the superficial, deep, and avascular plexuses were obtained for a total of 705 starting images. Images were augmented with standard intensity and geometric transforms, and regions from adjacent plexuses were programmatically combined to create synthetic 2-class images for each OCTA cube. Images were partitioned on a per patient basis into training, validation, and reserved test groups to train and evaluate a U-Net based machine learning model. To investigate the generalization of the model, we applied the model to multiclass thin slabs from OCTA volumes and qualitatively observed the resulting b-scans.</div></div><div><h3>Main Outcome Measures</h3><div>Plexus segmentation performance was assessed quantitatively using Dice scores on a held-out test set.</div></div><div><h3>Results</h3><div>After training on single-class plexus images, our model achieved good results (Dice scores &gt; 0.82) and was further improved when using the synthetic 2-class images (Dice &gt;0.95). Although not trained on more complex multiclass slabs, the model performed plexus labeling on slab data, which indicates that the use of only OCTA data shows promise for segmenting the superficial, deep, and avascular plexuses without requiring OCT layer segmentations, and the use of synthetic 2-class images makes a significant performance improvement.</div></div><div><h3>Conclusions</h3><div>This study presents the use of OCTA data alone to segment the superficial, deep, and avascular plexuses of the retina, confirming that use of structural OCT layer segmentations as boundaries is not required.</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 1","pages":"Article 100605"},"PeriodicalIF":3.2,"publicationDate":"2024-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142655080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Sources of Discrepancy between Retinal Nerve Fiber Layer and Bruch’s Membrane Opening-Minimum Rim Width Thickness in Eyes with Glaucoma","authors":"Iris Zhuang MD,&nbsp;Maryam Ashrafkhorasani MD,&nbsp;Vahid Mohammadzadeh MD,&nbsp;Kouros Nouri-Mahdavi MD, MS","doi":"10.1016/j.xops.2024.100601","DOIUrl":"10.1016/j.xops.2024.100601","url":null,"abstract":"<div><h3>Purpose</h3><div>To compare the discrepancies between circumpapillary retinal nerve fiber layer (RNFL) and Bruch’s membrane opening-minimum rim width (BMO-MRW) thickness in glaucoma eyes.</div></div><div><h3>Design</h3><div>A cross-sectional observational study.</div></div><div><h3>Subjects</h3><div>One hundred eighty-six eyes (118 patients) with glaucoma.</div></div><div><h3>Methods</h3><div>OCT optic nerve head volume scans of patients enrolled in the Advanced Glaucoma Progression Study at the final available visit were exported. The RNFL and BMO-MRW measurements were averaged into corresponding 7.5° sectors, and the nasal sector data were excluded from analyses. A 2-stage screening process was used to identify true mismatches between the RNFL and BMO-MRW measurements, in which either the RNFL or BMO-MRW value was in the less than first percentile range while its counterpart was in the greater than first percentile range on the temporal-superior-nasal-inferior-temporal curve. The prevalence of these mismatches was mapped, and corresponding images were reviewed to determine the underlying cause of these discrepancies.</div></div><div><h3>Main Outcome Measures</h3><div>Proportion of mismatches between RNFL and BMO-MRW, location of mismatches between RNFL and BMO-MRW, anatomical causes of mismatches between RNFL and BMO-MRW.</div></div><div><h3>Results</h3><div>Mismatch analysis revealed true mismatches between RNFL and BMO-MRW in 7.7% of sectors. High BMO-MRW with low corresponding RNFL mismatches were most frequently located at the 45° and 322.5° sectors, whereas high RNFL with corresponding low BMO-MRW mismatches peaked at the 75° sector. Large blood vessels accounted for 90.9% of high RNFL with low BMO-MRW mismatches. Small to large blood vessels accounted for 62.9% of high BMO-MRW with low RNFL mismatches; the remaining mismatches could be attributed to retinoschisis or inclusion of outer retinal layers in BMO-MRW measurements.</div></div><div><h3>Conclusions</h3><div>Although overall agreement between RNFL and BMO-MRW measurements is good in areas with advanced damage, blood vessels and other anatomical factors can cause discrepancies between the 2 types of structural measurements and need to be considered when evaluating the utility of such measurements for detection of change.</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 1","pages":"Article 100601"},"PeriodicalIF":3.2,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142423212","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diagnosing Glaucoma Based on the Ocular Hypertension Treatment Study Dataset Using Chat Generative Pre-Trained Transformer as a Large Language Model","authors":"Hina Raja PhD ,&nbsp;Xiaoqin Huang PhD ,&nbsp;Mohammad Delsoz MD ,&nbsp;Yeganeh Madadi PhD ,&nbsp;Asma Poursoroush PhD ,&nbsp;Asim Munawar PhD ,&nbsp;Malik Y. Kahook MD ,&nbsp;Siamak Yousefi PhD","doi":"10.1016/j.xops.2024.100599","DOIUrl":"10.1016/j.xops.2024.100599","url":null,"abstract":"<div><h3>Purpose</h3><p>To evaluate the capabilities of Chat Generative Pre-Trained Transformer (ChatGPT), as a large language model (LLM), for diagnosing glaucoma using the Ocular Hypertension Treatment Study (OHTS) dataset, and comparing the diagnostic capability of ChatGPT 3.5 and ChatGPT 4.0.</p></div><div><h3>Design</h3><p>Prospective data collection study.</p></div><div><h3>Participants</h3><p>A total of 3170 eyes of 1585 subjects from the OHTS were included in this study.</p></div><div><h3>Methods</h3><p>We selected demographic, clinical, ocular, visual field, optic nerve head photo, and history of disease parameters of each participant and developed case reports by converting tabular data into textual format based on information from both eyes of all subjects. We then developed a procedure using the application programming interface of ChatGPT, a LLM-based chatbot, to automatically input prompts into a chat box. This was followed by querying 2 different generations of ChatGPT (versions 3.5 and 4.0) regarding the underlying diagnosis of each subject. We then evaluated the output responses based on several objective metrics.</p></div><div><h3>Main Outcome Measures</h3><p>Area under the receiver operating characteristic curve (AUC), accuracy, specificity, sensitivity, and F1 score.</p></div><div><h3>Results</h3><p>Chat Generative Pre-Trained Transformer 3.5 achieved AUC of 0.74, accuracy of 66%, specificity of 64%, sensitivity of 85%, and F1 score of 0.72. Chat Generative Pre-Trained Transformer 4.0 obtained AUC of 0.76, accuracy of 87%, specificity of 90%, sensitivity of 61%, and F1 score of 0.92.</p></div><div><h3>Conclusions</h3><p>The accuracy of ChatGPT 4.0 in diagnosing glaucoma based on input data from OHTS was promising. The overall accuracy of ChatGPT 4.0 was higher than ChatGPT 3.5. However, ChatGPT 3.5 was found to be more sensitive than ChatGPT 4.0. In its current forms, ChatGPT may serve as a useful tool in exploring disease status of ocular hypertensive eyes when specific data are available for analysis. In the future, leveraging LLMs with multimodal capabilities, allowing for integration of imaging and diagnostic testing as part of the analyses, could further enhance diagnostic capabilities and enhance diagnostic accuracy.</p></div><div><h3>Financial Disclosures</h3><p>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</p></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 1","pages":"Article 100599"},"PeriodicalIF":3.2,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001350/pdfft?md5=9446852d5e50ba948a58b4ce06421174&pid=1-s2.0-S2666914524001350-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142270457","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Computational Framework for Intraoperative Pupil Analysis in Cataract Surgery","authors":"Binh Duong Giap PhD ,&nbsp;Karthik Srinivasan MD, MS ,&nbsp;Ossama Mahmoud MD ,&nbsp;Dena Ballouz MD ,&nbsp;Jefferson Lustre BS ,&nbsp;Keely Likosky BS ,&nbsp;Shahzad I. Mian MD ,&nbsp;Bradford L. Tannen MD, JD ,&nbsp;Nambi Nallasamy MD","doi":"10.1016/j.xops.2024.100597","DOIUrl":"10.1016/j.xops.2024.100597","url":null,"abstract":"<div><h3>Purpose</h3><div>Pupillary instability is a known risk factor for complications in cataract surgery. This study aims to develop and validate an innovative and reliable computational framework for the automated assessment of pupil morphologic changes during the various phases of cataract surgery.</div></div><div><h3>Design</h3><div>Retrospective surgical video analysis.</div></div><div><h3>Subjects</h3><div>Two hundred forty complete surgical video recordings, among which 190 surgeries were conducted without the use of pupil expansion devices (PEDs) and 50 were performed with the use of a PED.</div></div><div><h3>Methods</h3><div>The proposed framework consists of 3 stages: feature extraction, deep learning (DL)-based anatomy recognition, and obstruction (OB) detection/compensation. In the first stage, surgical video frames undergo noise reduction using a tensor-based wavelet feature extraction method. In the second stage, DL-based segmentation models are trained and employed to segment the pupil, limbus, and palpebral fissure. In the third stage, obstructed visualization of the pupil is detected and compensated for using a DL-based algorithm. A dataset of 5700 intraoperative video frames across 190 cataract surgeries in the BigCat database was collected for validating algorithm performance.</div></div><div><h3>Main Outcome Measures</h3><div>The pupil analysis framework was assessed on the basis of segmentation performance for both obstructed and unobstructed pupils. Classification performance of models utilizing the segmented pupil time series to predict surgeon use of a PED was also assessed.</div></div><div><h3>Results</h3><div>An architecture based on the Feature Pyramid Network model with Visual Geometry Group 16 backbone integrated with the adaptive wavelet tensor feature extraction feature extraction method demonstrated the highest performance in anatomy segmentation, with Dice coefficient of 96.52%. Incorporation of an OB compensation algorithm improved performance further (Dice 96.82%). Downstream analysis of framework output enabled the development of a Support Vector Machine–based classifier that could predict surgeon usage of a PED prior to its placement with 96.67% accuracy and area under the curve of 99.44%.</div></div><div><h3>Conclusions</h3><div>The experimental results demonstrate that the proposed framework (1) provides high accuracy in pupil analysis compared with human-annotated ground truth, (2) substantially outperforms isolated use of a DL segmentation model, and (3) can enable downstream analytics with clinically valuable predictive capacity.</div></div><div><h3>Financial Disclosures</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 1","pages":"Article 100597"},"PeriodicalIF":3.2,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142423213","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Relationship between Neighborhood-Level Social Risk Factor Measures and Presenting Glaucoma Severity Utilizing Multilevel Modeling","authors":"Patrice M. Hicks PhD, MPH ,&nbsp;Ming-Chen Lu MS ,&nbsp;Maria A. Woodward MD, MS ,&nbsp;Leslie M. Niziol MS ,&nbsp;Deborah Darnley-Fisch MD ,&nbsp;Michele Heisler MD ,&nbsp;Kenneth Resnicow PhD ,&nbsp;David C. Musch PhD, MPH ,&nbsp;Jamie Mitchell PhD, MSW ,&nbsp;Roshanak Mehdipanah PhD, MS ,&nbsp;Nauman R. Imami MD ,&nbsp;Paula Anne Newman-Casey MD MS","doi":"10.1016/j.xops.2024.100598","DOIUrl":"10.1016/j.xops.2024.100598","url":null,"abstract":"<div><h3>Purpose</h3><p>The neighborhood and built environment social determinant of health domain has several social risk factors (SRFs) that are modifiable through policy efforts. We investigated the impact of neighborhood-level SRFs on presenting glaucoma severity at a tertiary eye care center.</p></div><div><h3>Design</h3><p>A cross-sectional study from August 2012 to May 2022 in the University of Michigan electronic health record (EHR).</p></div><div><h3>Participants</h3><p>Patients with a diagnosis of any open-angle glaucoma with ≥1 eye care visit at the University of Michigan Kellogg Eye Center and ≥1 reliable visual field (VF).</p></div><div><h3>Methods</h3><p>Participants who met inclusion criteria were identified by International Classification of Diseases ninth and tenth revision codes (365.x/H40.x). Data extracted from the EHR included patient demographics, address, presenting mean deviation (MD), and VF reliability. Addresses were mapped to SRF measures at the census tract, block group, and county levels. Multilevel linear regression models were used to estimate the fixed effects of each SRF on MD, after adjusting for patient-level demographic factors and a random effect for neighborhood. Interactions between each SRF measure with patient-level race and Medicaid status were tested for an additive effect on MD.</p></div><div><h3>Main Outcome Measures</h3><p>The main outcome measure was the effect of SRF on presenting MD.</p></div><div><h3>Results</h3><p>In total, 4428 patients were included in the analysis who were, on average, 70.3 years old (standard deviation = 11.9), 52.6% self-identified as female, 75.8% self-identified as White race, and 8.9% had Medicaid. The median value of presenting MD was −4.94 decibels (dB) (interquartile range = −11.45 to −2.07 dB). Neighborhood differences accounted for 4.4% of the variability in presenting MD. Neighborhood-level measures, including worse area deprivation (estimate, β = −0.31 per 1-unit increase; <em>P</em> &lt; 0.001), increased segregation (β = −0.92 per 0.1-unit increase in Theil’s H index; <em>P</em> &lt; 0.001), and increased neighborhood Medicaid (β = −0.68; <em>P</em> &lt; 0.001) were associated with worse presenting MD. Significant interaction effects with race and Medicaid status were found in several neighborhood-level SRF measures.</p></div><div><h3>Conclusions</h3><p>Although patients’ neighborhood SRF measures accounted for a minority of the variability in presenting MD, most neighborhood-level SRFs are modifiable and were associated with clinically meaningful differences in presenting MD. Policies that aim to reduce neighborhood inequities by addressing allocation of resources could have lasting impacts on vision outcomes.</p></div><div><h3>Financial Disclosure(s)</h3><p>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</p></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 1","pages":"Article 100598"},"PeriodicalIF":3.2,"publicationDate":"2024-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001349/pdfft?md5=241edf796058dadff7c3bb9a45c9e13d&pid=1-s2.0-S2666914524001349-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142270455","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Impact of Race, Ethnicity, and Sex on Fairness in Artificial Intelligence for Glaucoma Prediction Models","authors":"Rohith Ravindranath MS ,&nbsp;Joshua D. Stein MD, MS ,&nbsp;Tina Hernandez-Boussard ,&nbsp;A. Caroline Fisher ,&nbsp;Sophia Y. Wang MD, MS","doi":"10.1016/j.xops.2024.100596","DOIUrl":"10.1016/j.xops.2024.100596","url":null,"abstract":"<div><h3>Objective</h3><div>Despite advances in artificial intelligence (AI) in glaucoma prediction, most works lack multicenter focus and do not consider fairness concerning sex, race, or ethnicity. This study aims to examine the impact of these sensitive attributes on developing fair AI models that predict glaucoma progression to necessitating incisional glaucoma surgery.</div></div><div><h3>Design</h3><div>Database study.</div></div><div><h3>Participants</h3><div>Thirty-nine thousand ninety patients with glaucoma, as identified by International Classification of Disease codes from 7 academic eye centers participating in the Sight OUtcomes Research Collaborative.</div></div><div><h3>Methods</h3><div>We developed XGBoost models using 3 approaches: (1) excluding sensitive attributes as input features, (2) including them explicitly as input features, and (3) training separate models for each group. Model input features included demographic details, diagnosis codes, medications, and clinical information (intraocular pressure, visual acuity, etc.), from electronic health records. The models were trained on patients from 5 sites (N = 27 999) and evaluated on a held-out internal test set (N = 3499) and 2 external test sets consisting of N = 1550 and N = 2542 patients.</div></div><div><h3>Main Outcomes and Measures</h3><div>Area under the receiver operating characteristic curve (AUROC) and equalized odds on the test set and external sites.</div></div><div><h3>Results</h3><div>Six thousand six hundred eighty-two (17.1%) of 39 090 patients underwent glaucoma surgery with a mean age of 70.1 (standard deviation 14.6) years, 54.5% female, 62.3% White, 22.1% Black, and 4.7% Latinx/Hispanic. We found that not including the sensitive attributes led to better classification performance (AUROC: 0.77–0.82) but worsened fairness when evaluated on the internal test set. However, on external test sites, the opposite was true: including sensitive attributes resulted in better classification performance (AUROC: external #1 - [0.73–0.81], external #2 - [0.67–0.70]), but varying degrees of fairness for sex and race as measured by equalized odds.</div></div><div><h3>Conclusions</h3><div>Artificial intelligence models predicting whether patients with glaucoma progress to surgery demonstrated bias with respect to sex, race, and ethnicity. The effect of sensitive attribute inclusion and exclusion on fairness and performance varied based on internal versus external test sets. Prior to deployment, AI models should be evaluated for fairness on the target population.</div></div><div><h3>Financial Disclosures</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 1","pages":"Article 100596"},"PeriodicalIF":3.2,"publicationDate":"2024-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001325/pdfft?md5=a7947c05f20d148756a130892f021b56&pid=1-s2.0-S2666914524001325-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142311980","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"XOLARIS: A 24-Month, Prospective, Natural History Study of 201 Participants with Retinitis Pigmentosa GTPase Regulator-Associated X-Linked Retinitis Pigmentosa","authors":"Robert E. MacLaren DPhil, FACS ,&nbsp;Jacque L. Duncan MD ,&nbsp;M. Dominik Fischer MD, DPhil ,&nbsp;Byron L. Lam MD ,&nbsp;Isabelle Meunier MD, PhD ,&nbsp;Mark E. Pennesi MD, PhD ,&nbsp;Eeva-Marja K. Sankila MD, PhD ,&nbsp;James A. Gow MD, MBA ,&nbsp;Jiang Li MS, MA ,&nbsp;So-Fai Tsang MD","doi":"10.1016/j.xops.2024.100595","DOIUrl":"10.1016/j.xops.2024.100595","url":null,"abstract":"<div><h3>Objective</h3><div>To improve the understanding of the natural disease progression of <em>retinitis pigmentosa GTPase</em> <em>regulator</em> (<em>RPGR</em>)<em>-</em>associated X-linked retinitis pigmentosa (XLRP).</div></div><div><h3>Design</h3><div>A multicenter, prospective, observational natural history study over 24 months.</div></div><div><h3>Participants</h3><div>Male participants aged ≥7 years with a pathogenic variant in the <em>RPGR</em> gene, a best-corrected visual acuity (BCVA) score of ≥34 ETDRS letters, and a mean 68-loci retinal sensitivity (assessed by microperimetry) of 0.1 to 20 decibels (dB).</div></div><div><h3>Methods</h3><div>Participants were divided into subgroups based on their BCVA score at baseline: 34 to 73 (lower BCVA) or ≥74 (higher BCVA) ETDRS letters. There were 7 visits over 24 months.</div></div><div><h3>Main Outcome Measures</h3><div>Change from baseline in BCVA, retinal sensitivity, low luminance visual acuity (LLVA), fixation stability, contrast sensitivity, visual field, anatomical measures, 25-item Visual Function Questionnaire (VFQ-25), intraocular pressure, and adverse events (AEs).</div></div><div><h3>Results</h3><div>Overall, 201 participants were included. The mean (standard deviation [SD]) age was 30.3 (11.9) years in the lower BCVA subgroup (n = 170) and 27.7 (10.1) years in the higher BCVA subgroup (n = 31). The study eye baseline mean (SD) BCVA scores were 59.4 (10.30) and 77.3 (3.95) in the lower and higher BCVA subgroups, respectively; the lower BCVA subgroup had lower retinal sensitivity in the study eye at baseline than the higher BCVA subgroup. Over 24 months, there were small observed changes in BCVA, retinal sensitivity, LLVA, fixation, contrast sensitivity, and fundus photography findings. There were observed mean (SD) changes at 24 months in the lower and higher BCVA subgroups of −1.01 (4.67) and 0.03 (5.83) dB-steradians in the volume of full-field hill of vision, −330.6 (869.51) and −122.7 (22.01) μm in distance from foveal center to the nearest border of preserved fundus autofluorescence, −104.3 (277.80) and −207.1 (171.01) μm in central ellipsoid width, and −2.8 (9.7) and −0.6 (7.6) in VFQ-25 composite score, respectively. There was 1 death from completed suicide. There were no ocular serious adverse events, and most AEs were mild/moderate.</div></div><div><h3>Conclusions</h3><div>This study provides evidence of the slow natural progression of XLRP over 24 months in both subgroups and provides important functional, anatomical, and safety data.</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 1","pages":"Article 100595"},"PeriodicalIF":3.2,"publicationDate":"2024-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142442773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of the Appropriateness and Readability of ChatGPT-4 Responses to Patient Queries on Uveitis","authors":"S. Saeed Mohammadi MD ,&nbsp;Anadi Khatri MD ,&nbsp;Tanya Jain MBBS, DNB ,&nbsp;Zheng Xian Thng MD ,&nbsp;Woong-sun Yoo MD, PhD ,&nbsp;Negin Yavari MD ,&nbsp;Vahid Bazojoo MD ,&nbsp;Azadeh Mobasserian MD ,&nbsp;Amir Akhavanrezayat MD ,&nbsp;Ngoc Trong Tuong Than MD ,&nbsp;Osama Elaraby MD ,&nbsp;Battuya Ganbold MD ,&nbsp;Dalia El Feky MD ,&nbsp;Ba Trung Nguyen MD ,&nbsp;Cigdem Yasar MD ,&nbsp;Ankur Gupta MD, MS ,&nbsp;Jia-Horung Hung MD ,&nbsp;Quan Dong Nguyen MD, MSc","doi":"10.1016/j.xops.2024.100594","DOIUrl":"10.1016/j.xops.2024.100594","url":null,"abstract":"<div><h3>Purpose</h3><div>To compare the utility of ChatGPT-4 as an online uveitis patient education resource with existing patient education websites.</div></div><div><h3>Design</h3><div>Evaluation of technology.</div></div><div><h3>Participants</h3><div>Not applicable.</div></div><div><h3>Methods</h3><div>The term “uveitis” was entered into the Google search engine, and the first 8 nonsponsored websites were selected to be enrolled in the study. Information regarding uveitis for patients was extracted from Healthline, Mayo Clinic, WebMD, National Eye Institute, Ocular Uveitis and Immunology Foundation, American Academy of Ophthalmology, Cleveland Clinic, and National Health Service websites. ChatGPT-4 was then prompted to generate responses about uveitis in both standard and simplified formats. To generate the simplified response, the following request was added to the prompt: 'Please provide a response suitable for the average American adult, at a sixth-grade comprehension level.’ Three dual fellowship-trained specialists, all masked to the sources, graded the appropriateness of the contents (extracted from the existing websites) and responses (generated responses by ChatGPT-4) in terms of personal preference, comprehensiveness, and accuracy. Additionally, 5 readability indices, including Flesch Reading Ease, Flesch–Kincaid Grade Level, Gunning Fog Index, Coleman–Liau Index, and Simple Measure of Gobbledygook index were calculated using an online calculator, Readable.com, to assess the ease of comprehension of each answer.</div></div><div><h3>Main Outcome Measures</h3><div>Personal preference, accuracy, comprehensiveness, and readability of contents and responses about uveitis.</div></div><div><h3>Results</h3><div>A total of 497 contents and responses, including 71 contents from existing websites, 213 standard responses, and 213 simplified responses from ChatGPT-4 were recorded and graded. Standard ChatGPT-4 responses were preferred and perceived to be more comprehensive by dually trained (uveitis and retina) specialist ophthalmologists while maintaining similar accuracy level compared with existing websites. Moreover, simplified ChatGPT-4 responses matched almost all existing websites in terms of personal preference, accuracy, and comprehensiveness. Notably, almost all readability indices suggested that standard ChatGPT-4 responses demand a higher educational level for comprehension, whereas simplified responses required lower level of education compared with the existing websites.</div></div><div><h3>Conclusions</h3><div>This study shows that ChatGPT can provide patients with an avenue to access comprehensive and accurate information about uveitis, tailored to their educational level.</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 1","pages":"Article 100594"},"PeriodicalIF":3.2,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142423210","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparison between Spectral-Domain and Swept-Source OCT Angiography for the Measurement of Persistent Hypertransmission Defects in Age-Related Macular Degeneration","authors":"Gissel Herrera MD ,&nbsp;Mengxi Shen MD, PhD ,&nbsp;Omer Trivizki MD ,&nbsp;Jeremy Liu MD ,&nbsp;Yingying Shi MD ,&nbsp;Farhan E. Hiya MD ,&nbsp;Jianqing Li MD ,&nbsp;Yuxuan Cheng BS ,&nbsp;Jie Lu MD, MS ,&nbsp;Qinqin Zhang PhD ,&nbsp;Robert C. O’Brien PhD ,&nbsp;Giovanni Gregori PhD ,&nbsp;Ruikang K. Wang PhD ,&nbsp;Philip J. Rosenfeld MD, PhD","doi":"10.1016/j.xops.2024.100593","DOIUrl":"10.1016/j.xops.2024.100593","url":null,"abstract":"<div><h3>Purpose</h3><p>Spectral-domain OCT angiography (SD-OCTA) scans were tested in an algorithm developed for use with swept-source OCT angiography (SS-OCTA) scans to determine if SD-OCTA scans yielded similar results for the detection and measurement of persistent choroidal hypertransmission defects (hyperTDs).</p></div><div><h3>Design</h3><p>Retrospective study.</p></div><div><h3>Participants</h3><p>Forty pairs of scans from 32 patients with late-stage nonexudative age-related macular degeneration (AMD).</p></div><div><h3>Methods</h3><p>Patients underwent both SD-OCTA and SS-OCTA imaging at the same visit using the 6 × 6 mm OCTA scan patterns. Using a semiautomatic algorithm that helped with outlining the hyperTDs, 2 graders independently validated persistent hyperTDs, which are defined as having a greatest linear dimension ≥250 μm on the en face images generated using a slab extending from 64 to 400 μm beneath Bruch’s membrane. The number of lesions and square root (sqrt) total area of the hyperTDs were obtained from the algorithm using each imaging method.</p></div><div><h3>Main Outcome Measures</h3><p>The mean sqrt area measurements and the number of hyperTDs were compared.</p></div><div><h3>Results</h3><p>The number of lesions and sqrt total area of the hyperTDs were highly concordant between the 2 instruments (r_c = 0.969 and r_c = 0.999, respectively). The mean number of hyperTDs was 4.3 ± 3.1 for SD-OCTA scans and 4.5 ± 3.3 for SS-OCTA scans (<em>P</em> = 0.06). The mean sqrt total area measurements were 1.16 ± 0.64 mm for the SD-OCTA scans and 1.17 ± 0.65 mm for the SS-OCTA scans (<em>P</em> &lt; 0.001). Because of the small standard error of the differences, the mean difference between the scans was statistically significant but not clinically significant.</p></div><div><h3>Conclusions</h3><p>Spectral-domain OCTA scans provide similar results to SS-OCTA scans when used to obtain the number and area measurements of persistent hyperTDs through a semiautomated algorithm previously developed for SS-OCTA. This facilitates the detection of atrophy with a more widely available scan pattern and the longitudinal study of early to late-stage AMD.</p></div><div><h3>Financial Disclosure(s)</h3><p>Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.</p></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"5 1","pages":"Article 100593"},"PeriodicalIF":3.2,"publicationDate":"2024-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001295/pdfft?md5=1d219e52f1d3cce12729e047e485c32e&pid=1-s2.0-S2666914524001295-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142232580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparison of Machine Learning Models to a Novel Score in the Identification of Patients at Low Risk for Diabetic Retinopathy","authors":"Amanda Luong BS ,&nbsp;Jesse Cheung BS ,&nbsp;Shyla McMurtry MD ,&nbsp;Christina Nelson BS ,&nbsp;Tyler Najac MD ,&nbsp;Philippe Ortiz MD ,&nbsp;Stephen Aronoff MD, MBA ,&nbsp;Jeffrey Henderer MD ,&nbsp;Yi Zhang MD, PhD","doi":"10.1016/j.xops.2024.100592","DOIUrl":"10.1016/j.xops.2024.100592","url":null,"abstract":"<div><h3>Purpose</h3><div>To develop an easily applicable predictor of patients at low risk for diabetic retinopathy (DR).</div></div><div><h3>Design</h3><div>An experimental study on the development and validation of machine learning models (MLMs) and a novel retinopathy risk score (RRS) to detect patients at low risk for DR.</div></div><div><h3>Subjects</h3><div>All individuals aged ≥18 years of age who participated in the telemedicine retinal screening initiative through Temple University Health Systems from October 1, 2016 through December 31, 2020. The subjects must have documented evidence of their diabetes mellitus (DM) diagnosis as well as a documented glycosylated hemoglobin (HbA1c) recorded in their chart within 6 months of the retinal screening photograph.</div></div><div><h3>Methods</h3><div>The charts of 1930 subjects (1590 evaluable) undergoing telemedicine screening for DR were reviewed, and 30 demographic and clinical parameters were collected. Diabetic retinopathy is a dichotomous variable where low risk is defined as no or mild retinopathy using the International Clinical Diabetic Retinopathy severity score. Five MLMs were trained to predict patients at low risk for DR using 1050 subjects and further underwent 10-fold cross validation to maximize its performance indicated by the area under the receiver operator characteristic curve (AUC). Additionally, a novel RRS is defined as the product of HbA1c closest to screening and years with DM. Retinopathy risk score was also applied to generate a predictive model.</div></div><div><h3>Main Outcome Measures</h3><div>The performance of the trained MLMs and the RRS model was compared using DeLong’s test. The models were further validated using a separate unseen test set of 540 subjects. The performance of the validation models were compared using DeLong’s test and chi-square tests.</div></div><div><h3>Results</h3><div>Using the test set, the AUC for the RRS was not statistically different from 4 out of 5 MLM. The error rate for predicting low-risk patients using the RRS was significantly lower than the naive rate (0.097 vs. 0.19; <em>P</em> &lt; 0.0001), and it was comparable to the error rates of the MLMs.</div></div><div><h3>Conclusions</h3><div>This novel RRS is a potentially useful and easily deployable predictor of patients at low risk for DR.</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 1","pages":"Article 100592"},"PeriodicalIF":3.2,"publicationDate":"2024-08-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142322310","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}