Ophthalmology science最新文献

{"title":"Advancing Glaucoma Diagnosis: Employing Confidence-Calibrated Label Smoothing Loss for Model Calibration","authors":"Midhula Vijayan PhD,&nbsp;Deepthi Keshav Prasad PhD,&nbsp;Venkatakrishnan Srinivasan MTech","doi":"10.1016/j.xops.2024.100555","DOIUrl":"10.1016/j.xops.2024.100555","url":null,"abstract":"<div><h3>Objective</h3><p>The aim of our research is to enhance the calibration of machine learning models for glaucoma classification through a specialized loss function named Confidence-Calibrated Label Smoothing (CC-LS) loss. This approach is specifically designed to refine model calibration without compromising accuracy by integrating label smoothing and confidence penalty techniques, tailored to the specifics of glaucoma detection.</p></div><div><h3>Design</h3><p>This study focuses on the development and evaluation of a calibrated deep learning model.</p></div><div><h3>Participants</h3><p>The study employs fundus images from both external datasets—the Online Retinal Fundus Image Database for Glaucoma Analysis and Research (482 normal, 168 glaucoma) and the Retinal Fundus Glaucoma Challenge (720 normal, 80 glaucoma)—and an extensive internal dataset (4639 images per category), aiming to bolster the model's generalizability. The model's clinical performance is validated using a comprehensive test set (47 913 normal, 1629 glaucoma) from the internal dataset.</p></div><div><h3>Methods</h3><p>The CC-LS loss function seamlessly integrates label smoothing, which tempers extreme predictions to avoid overfitting, with confidence-based penalties. These penalties deter the model from expressing undue confidence in incorrect classifications. Our study aims at training models using the CC-LS and comparing their performance with those trained using conventional loss functions.</p></div><div><h3>Main Outcome Measures</h3><p>The model's precision is evaluated using metrics like the Brier score, sensitivity, specificity, and the false positive rate, alongside qualitative heatmap analyses for a holistic accuracy assessment.</p></div><div><h3>Results</h3><p>Preliminary findings reveal that models employing the CC-LS mechanism exhibit superior calibration metrics, as evidenced by a Brier score of 0.098, along with notable accuracy measures: sensitivity of 81%, specificity of 80%, and weighted accuracy of 80%. Importantly, these enhancements in calibration are achieved without sacrificing classification accuracy.</p></div><div><h3>Conclusions</h3><p>The CC-LS loss function presents a significant advancement in the pursuit of deploying machine learning models for glaucoma diagnosis. By improving calibration, the CC-LS ensures that clinicians can interpret and trust the predictive probabilities, making artificial intelligence-driven diagnostic tools more clinically viable. From a clinical standpoint, this heightened trust and interpretability can potentially lead to more timely and appropriate interventions, thereby optimizing patient outcomes and safety.</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":"4 6","pages":"Article 100555"},"PeriodicalIF":3.2,"publicationDate":"2024-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524000915/pdfft?md5=427dfb03f669c0bff9531a39938549c2&pid=1-s2.0-S2666914524000915-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142006395","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":"Assessment of Retinal Volume in Individuals Without Ocular Disorders Based on Wide-Field Swept-Source OCT","authors":"Yoshiaki Chiku MD,&nbsp;Takao Hirano MD, PhD,&nbsp;Ken Hoshiyama MD,&nbsp;Yasuhiro Iesato MD, PhD,&nbsp;Toshinori Murata MD, PhD","doi":"10.1016/j.xops.2024.100569","DOIUrl":"10.1016/j.xops.2024.100569","url":null,"abstract":"<div><h3>Purpose</h3><p>To evaluate retinal volume (RV) in eyes without retinal disease using wide-field swept-source OCT (SS-OCT).</p></div><div><h3>Design</h3><p>Observational, cross-sectional design.</p></div><div><h3>Participants</h3><p>A total of 332 eyes of 166 healthy participants.</p></div><div><h3>Methods</h3><p>The eyes were imaged with OCT-S1 (Canon) using a protocol centered on the fovea cube scans (20 × 23 mm) of SS-OCT images. Retinal volume (6-mm circle, 6–20-mm ring) and various parameters were evaluated in a multivariate analysis using a generalized estimating equation model. Each quadrant of the macula except for the fovea (1–6 mm in diameter) and peripheral ring (6–20 mm in diameter) was also evaluated.</p></div><div><h3>Main Outcome Measures</h3><p>Retinal volume.</p></div><div><h3>Results</h3><p>In the multivariate analysis, older age and longer axial length were associated with smaller macular RV, whereas older age and left eye were associated with smaller peripheral RV. The temporal area was significantly smaller than all other areas in the macula (1–6 mm), whereas the inferior area was significantly smaller than all other areas in the peripheral retina (6–20 mm).</p></div><div><h3>Conclusions</h3><p>In wide-field SS-OCT images, age and left eye are negatively correlated with peripheral RV. The thinnest part of the retinal quadrant differs between the macular and peripheral retinas.</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":"4 6","pages":"Article 100569"},"PeriodicalIF":3.2,"publicationDate":"2024-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001052/pdfft?md5=394b3468fb665e4d2094d378b0d55297&pid=1-s2.0-S2666914524001052-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142020946","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":"Artificial Intelligence to Facilitate Clinical Trial Recruitment in Age-Related Macular Degeneration","authors":"Dominic J. Williamson MSc ,&nbsp;Robbert R. Struyven MD ,&nbsp;Fares Antaki MD ,&nbsp;Mark A. Chia MD ,&nbsp;Siegfried K. Wagner MD, PhD ,&nbsp;Mahima Jhingan MBBS ,&nbsp;Zhichao Wu PhD ,&nbsp;Robyn Guymer MBBS, PhD ,&nbsp;Simon S. Skene PhD ,&nbsp;Naaman Tammuz PhD ,&nbsp;Blaise Thomson PhD ,&nbsp;Reena Chopra PhD ,&nbsp;Pearse A. Keane MD","doi":"10.1016/j.xops.2024.100566","DOIUrl":"10.1016/j.xops.2024.100566","url":null,"abstract":"<div><h3>Objective</h3><p>Recent developments in artificial intelligence (AI) have positioned it to transform several stages of the clinical trial process. In this study, we explore the role of AI in clinical trial recruitment of individuals with geographic atrophy (GA), an advanced stage of age-related macular degeneration, amidst numerous ongoing clinical trials for this condition.</p></div><div><h3>Design</h3><p>Cross-sectional study.</p></div><div><h3>Subjects</h3><p>Retrospective dataset from the INSIGHT Health Data Research Hub at Moorfields Eye Hospital in London, United Kingdom, including 306 651 patients (602 826 eyes) with suspected retinal disease who underwent OCT imaging between January 1, 2008 and April 10, 2023.</p></div><div><h3>Methods</h3><p>A deep learning model was trained on OCT scans to identify patients potentially eligible for GA trials, using AI-generated segmentations of retinal tissue. This method's efficacy was compared against a traditional keyword-based electronic health record (EHR) search. A clinical validation with fundus autofluorescence (FAF) images was performed to calculate the positive predictive value of this approach, by comparing AI predictions with expert assessments.</p></div><div><h3>Main Outcome Measures</h3><p>The primary outcomes included the positive predictive value of AI in identifying trial-eligible patients, and the secondary outcome was the intraclass correlation between GA areas segmented on FAF by experts and AI-segmented OCT scans.</p></div><div><h3>Results</h3><p>The AI system shortlisted a larger number of eligible patients with greater precision (1139, positive predictive value: 63%; 95% confidence interval [CI]: 54%–71%) compared with the EHR search (693, positive predictive value: 40%; 95% CI: 39%–42%). A combined AI-EHR approach identified 604 eligible patients with a positive predictive value of 86% (95% CI: 79%–92%). Intraclass correlation of GA area segmented on FAF versus AI-segmented area on OCT was 0.77 (95% CI: 0.68–0.84) for cases meeting trial criteria. The AI also adjusts to the distinct imaging criteria from several clinical trials, generating tailored shortlists ranging from 438 to 1817 patients.</p></div><div><h3>Conclusions</h3><p>This study demonstrates the potential for AI in facilitating automated prescreening for clinical trials in GA, enabling site feasibility assessments, data-driven protocol design, and cost reduction. Once treatments are available, similar AI systems could also be used to identify individuals who may benefit from treatment.</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":"4 6","pages":"Article 100566"},"PeriodicalIF":3.2,"publicationDate":"2024-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001027/pdfft?md5=d3fd08da8ceda3860a8f32aa944ff57b&pid=1-s2.0-S2666914524001027-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141949722","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":"Artificial Intelligence-Based Disease Activity Monitoring to Personalized Neovascular Age-Related Macular Degeneration Treatment: A Feasibility Study","authors":"Zufar Mulyukov PhD ,&nbsp;Pearse A. Keane FRCOphth, MD ,&nbsp;Jayashree Sahni FRCOphth, MD ,&nbsp;Sandra Liakopoulos MD ,&nbsp;Katja Hatz MD ,&nbsp;Daniel Shu Wei Ting MD, PhD ,&nbsp;Roberto Gallego-Pinazo MD, PhD ,&nbsp;Tariq Aslam PhD, DM(Oxon) ,&nbsp;Chui Ming Gemmy Cheung FRCOphth, MD ,&nbsp;Gabriella De Salvo FRCOphth, MD ,&nbsp;Oudy Semoun MD ,&nbsp;Gábor Márk Somfai MD, PhD ,&nbsp;Andreas Stahl MD ,&nbsp;Brandon J. Lujan MD ,&nbsp;Daniel Lorand MSc","doi":"10.1016/j.xops.2024.100565","DOIUrl":"10.1016/j.xops.2024.100565","url":null,"abstract":"<div><h3>Purpose</h3><p>To evaluate the performance of a disease activity (DA) model developed to detect DA in participants with neovascular age-related macular degeneration (nAMD).</p></div><div><h3>Design</h3><p>Post hoc analysis.</p></div><div><h3>Participants</h3><p>Patient dataset from the phase III HAWK and HARRIER (H&amp;H) studies.</p></div><div><h3>Methods</h3><p>An artificial intelligence (AI)-based DA model was developed to generate a DA score based on measurements of OCT images and other parameters collected from H&amp;H study participants. Disease activity assessments were classified into 3 categories based on the extent of agreement between the DA model’s scores and the H&amp;H investigators’ decisions: agreement (“easy”), disagreement (“noisy”), and close to the decision boundary (“difficult”). Then, a panel of 10 international retina specialists (“panelists”) reviewed a sample of DA assessments of these 3 categories that contributed to the training of the final DA model. A panelists’ majority vote on the reviewed cases was used to evaluate the accuracy, sensitivity, and specificity of the DA model.</p></div><div><h3>Main Outcome Measures</h3><p>The DA model’s performance in detecting DA compared with the DA assessments made by the investigators and panelists’ majority vote.</p></div><div><h3>Results</h3><p>A total of 4472 OCT DA assessments were used to develop the model; of these, panelists reviewed 425, categorized as “easy” (17.2%), “noisy” (20.5%), and “difficult” (62.4%). False-positive and false negative rates of the DA model’s assessments decreased after changing the assessment in some cases reviewed by the panelists and retraining the DA model. Overall, the DA model achieved 80% accuracy. For “easy” cases, the DA model reached 96% accuracy and performed as well as the investigators (96% accuracy) and panelists (90% accuracy). For “noisy” cases, the DA model performed similarly to panelists and outperformed the investigators (84%, 86%, and 16% accuracies, respectively). The DA model also outperformed the investigators for “difficult” cases (74% and 53% accuracies, respectively) but underperformed the panelists (86% accuracy) owing to lower specificity. Subretinal and intraretinal fluids were the main clinical parameters driving the DA assessments made by the panelists.</p></div><div><h3>Conclusions</h3><p>These results demonstrate the potential of using an AI-based DA model to optimize treatment decisions in the clinical setting and in detecting and monitoring DA in patients with nAMD.</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":"4 6","pages":"Article 100565"},"PeriodicalIF":3.2,"publicationDate":"2024-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001015/pdfft?md5=65ab512316012a4b8940a3501fab3963&pid=1-s2.0-S2666914524001015-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141998625","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":"Characterization of Retinal Microvascular Abnormalities in Birdshot Chorioretinopathy Using OCT Angiography","authors":"Aman Kumar MD ,&nbsp;Alexander Zeleny MD ,&nbsp;Sunil Bellur MD ,&nbsp;Natasha Kesav MD ,&nbsp;Enny Oyeniran MD ,&nbsp;Kübra Gul Olke MD ,&nbsp;Susan Vitale PhD, MHS ,&nbsp;Wijak Kongwattananon MD ,&nbsp;H. Nida Sen MD, MHS ,&nbsp;Shilpa Kodati MD","doi":"10.1016/j.xops.2024.100559","DOIUrl":"10.1016/j.xops.2024.100559","url":null,"abstract":"<div><h3>Objective</h3><p>To characterize changes in the retinal microvasculature in eyes with birdshot chorioretinopathy (BCR) using OCT angiography (OCTA).</p></div><div><h3>Design</h3><p>Retrospective, observational, single center.</p></div><div><h3>Subjects</h3><p>Twenty-eight patients (53 eyes) with BCR and 59 age-matched controls (110 eyes).</p></div><div><h3>Methods</h3><p>En face OCTA images of the superficial capillary plexus (SCP) and deep capillary plexus (DCP) of each eye were assessed for the presence of microvascular abnormalities and used to measure the vessel and foveal avascular zone (FAZ) areas. A longitudinal analysis was performed with a representative cohort of 23 BCR eyes (16 patients) at baseline and at a 2-year time point.</p></div><div><h3>Main Outcome Measures</h3><p>Whole-image vessel density (VD, %), extrafoveal avascular zone (extra-FAZ) VD (%), and FAZ area (%) were calculated and compared between control and BCR eyes. The frequency of microvascular abnormalities in BCR eyes was recorded.</p></div><div><h3>Results</h3><p>In the SCP, increased intercapillary space and capillary loops were common features present on OCTA images. Whole-image and extra-FAZ VD were lower in the BCR group compared with controls (<em>P</em> &lt; 0.0001 [SCP and DCP]). Foveal avascular zone area was enlarged in BCR eyes (<em>P</em> = 0.0008 [DCP]). Worsening best-corrected visual acuity was associated with a decrease in whole-image and extra-FAZ VD in the SCP (<em>P</em> &lt; 0.0001 for both) and the DCP (<em>P</em> &lt; 0.005 for both). Multivariable analysis, with vessel analysis parameters as outcomes, demonstrated that increasing age, increasing disease duration, lower central subfield thickness, and treatment-naive eyes (compared with those on only biologics) were associated with a significant decrease in both DCP whole-image and extra-FAZ VD. Increasing disease duration was associated with a significant decrease in both SCP whole-image and extra-FAZ VD. Longitudinal analysis demonstrated no significant difference in any vessel analysis parameters except for an increase in DCP FAZ area.</p></div><div><h3>Conclusions</h3><p>Our results demonstrate a significant a decrease in VD in BCR eyes and an association on multivariable analysis with disease duration. Quantifying VD in the retinal microvasculature may be a useful biomarker for monitoring disease severity and progression in patients with BCR. Further studies with extended longitudinal follow-up are needed to characterize its utility in monitoring disease progression and treatment response.</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":"4 6","pages":"Article 100559"},"PeriodicalIF":3.2,"publicationDate":"2024-06-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524000952/pdfft?md5=082f883617d8809ef87350cdecb8e690&pid=1-s2.0-S2666914524000952-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141961193","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 Diagnosis Codes to Clinical Notes in Classifying Patients with Diabetic Retinopathy","authors":"","doi":"10.1016/j.xops.2024.100564","DOIUrl":"10.1016/j.xops.2024.100564","url":null,"abstract":"<div><h3>Purpose</h3><p>Electronic health records (EHRs) contain a vast amount of clinical data. Improved automated classification approaches have the potential to accurately and efficiently identify patient cohorts for research. We evaluated if a rule-based natural language processing (NLP) algorithm using clinical notes performed better for classifying proliferative diabetic retinopathy (PDR) and nonproliferative diabetic retinopathy (NPDR) severity compared with International Classification of Diseases, ninth edition (ICD-9) or 10th edition (ICD-10) codes.</p></div><div><h3>Design</h3><p>Cross-sectional study.</p></div><div><h3>Subjects</h3><p>Deidentified EHR data from an academic medical center identified 2366 patients aged ≥18 years, with diabetes mellitus, diabetic retinopathy (DR), and available clinical notes.</p></div><div><h3>Methods</h3><p>From these 2366 patients, 306 random patients (100 training set, 206 test set) underwent chart review by ophthalmologists to establish the gold standard. International Classification of Diseases codes were extracted from the EHR. The notes algorithm identified positive mention of PDR and NPDR severity from clinical notes. Proliferative diabetic retinopathy and NPDR severity classification by ICD codes and the notes algorithm were compared with the gold standard. The entire DR cohort (N = 2366) was then classified as having presence (or absence) of PDR using ICD codes and the notes algorithm.</p></div><div><h3>Main Outcome Measures</h3><p>Sensitivity, specificity, positive predictive value (PPV), negative predictive value, and F1 score for the notes algorithm compared with ICD codes using a gold standard of chart review.</p></div><div><h3>Results</h3><p>For PDR classification of the test set patients, the notes algorithm performed better than ICD codes for all metrics. Specifically, the notes algorithm had significantly higher sensitivity (90.5% [95% confidence interval 85.7, 94.9] vs. 68.4% [60.4, 75.3]), but similar PPV (98.0% [95.4–100] vs. 94.7% [90.3, 98.3]) respectively. The F1 score was 0.941 [0.910, 0.966] for the notes algorithm compared with 0.794 [0.734, 0.842] for ICD codes. For PDR classification, ICD-10 codes performed better than ICD-9 codes (F1 score 0.836 [0.771, 0.878] vs. 0.596 [0.222, 0.692]). For NPDR severity classification, the notes algorithm performed similarly to ICD codes, but performance was limited by small sample size.</p></div><div><h3>Conclusions</h3><p>The notes algorithm outperformed ICD codes for PDR classification. The findings demonstrate the significant potential of applying a rule-based NLP algorithm to clinical notes to increase the efficiency and accuracy of cohort selection for research.</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":"4 6","pages":"Article 100564"},"PeriodicalIF":3.2,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524001003/pdfft?md5=958f568c39babd1a1573d7125a9a1d48&pid=1-s2.0-S2666914524001003-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141414826","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 Novel Time-Aware Deep Learning Model Predicting Myopia in Children and Adolescents","authors":"","doi":"10.1016/j.xops.2024.100563","DOIUrl":"10.1016/j.xops.2024.100563","url":null,"abstract":"<div><h3>Objective</h3><p>To quantitatively predict children’s and adolescents’ spherical equivalent (SE) by leveraging their variable-length historical vision records.</p></div><div><h3>Design</h3><p>Retrospective analysis.</p></div><div><h3>Participants</h3><p>Eight hundred ninety-five myopic children and adolescents aged 4 to 18 years, with a complete ophthalmic examination and retinoscopy in cycloplegia prior to spectacle correction, were enrolled in the period from January 1, 2008 to July 1, 2023 at the University Hospital “Sveti Duh,” Zagreb, Croatia.</p></div><div><h3>Methods</h3><p>A novel modification of time-aware long short-term memory (LSTM) was used to quantitatively predict children’s and adolescents’ SE within 7 years after diagnosis.</p></div><div><h3>Main Outcome Measures</h3><p>The utilization of extended gate time-aware LSTM involved capturing temporal features within irregularly sampled time series data. This approach aligned more closely with the characteristics of fact-based data, increasing its applicability and contributing to the early identification of myopia progression.</p></div><div><h3>Results</h3><p>The testing set exhibited a mean absolute prediction error (MAE) of 0.10 ± 0.15 diopter (D) for SE. Lower MAE values were associated with longer sequence lengths, shorter prediction durations, older age groups, and low myopia, while higher MAE values were observed with shorter sequence lengths, longer prediction durations, younger age groups, and in premyopic or high myopic individuals, ranging from as low as 0.03 ± 0.04 D to as high as 0.45 ± 0.24 D.</p></div><div><h3>Conclusions</h3><p>Extended gate time-aware LSTM capturing temporal features in irregularly sampled time series data can be used to quantitatively predict children’s and adolescents’ SE within 7 years with an overall error of 0.10 ± 0.15 D. This value is substantially lower than the threshold for prediction to be considered clinically acceptable, such as a criterion of 0.75 D.</p></div><div><h3>Financial Disclosure(s)</h3><p>The author(s) have no proprietary or commercial interest in any materials discussed in this article.</p></div>","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"4 6","pages":"Article 100563"},"PeriodicalIF":3.2,"publicationDate":"2024-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S266691452400099X/pdfft?md5=88ef2e36f9b015b3320de2e4a1942b13&pid=1-s2.0-S266691452400099X-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141396908","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":"Ocular Manifestations of Enterovirus: An Important Emerging Pathogen","authors":"","doi":"10.1016/j.xops.2024.100562","DOIUrl":"10.1016/j.xops.2024.100562","url":null,"abstract":"","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"4 6","pages":"Article 100562"},"PeriodicalIF":3.2,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524000988/pdfft?md5=6bc73bed8491ebd6df500f1082e1578f&pid=1-s2.0-S2666914524000988-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141280811","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":"Deep Learning–Based Clustering of OCT Images for Biomarker Discovery in Age-Related Macular Degeneration (PINNACLE Study Report 4)","authors":"Robbie Holland MEng ,&nbsp;Rebecca Kaye MD ,&nbsp;Ahmed M. Hagag MD ,&nbsp;Oliver Leingang PhD ,&nbsp;Thomas R.P. Taylor MD ,&nbsp;Hrvoje Bogunović PhD ,&nbsp;Ursula Schmidt-Erfurth MD ,&nbsp;Hendrik P.N. Scholl MD ,&nbsp;Daniel Rueckert PhD ,&nbsp;Andrew J. Lotery MD ,&nbsp;Sobha Sivaprasad MD ,&nbsp;Martin J. Menten PhD","doi":"10.1016/j.xops.2024.100543","DOIUrl":"10.1016/j.xops.2024.100543","url":null,"abstract":"<div><h3>Purpose</h3><p>We introduce a deep learning–based biomarker proposal system for the purpose of accelerating biomarker discovery in age-related macular degeneration (AMD).</p></div><div><h3>Design</h3><p>Retrospective analysis of a large data set of retinal OCT images.</p></div><div><h3>Participants</h3><p>A total of 3456 adults aged between 51 and 102 years whose OCT images were collected under the PINNACLE project.</p></div><div><h3>Methods</h3><p>Our system proposes candidates for novel AMD imaging biomarkers in OCT. It works by first training a neural network using self-supervised contrastive learning to discover, without any clinical annotations, features relating to both known and unknown AMD biomarkers present in 46 496 retinal OCT images. To interpret the learned biomarkers, we partition the images into 30 subsets, termed clusters, that contain similar features. We conduct 2 parallel 1.5-hour semistructured interviews with 2 independent teams of retinal specialists to assign descriptions in clinical language to each cluster. Descriptions of clusters achieving consensus can potentially inform new biomarker candidates.</p></div><div><h3>Main Outcome Measures</h3><p>We checked if each cluster showed clear features comprehensible to retinal specialists, if they related to AMD, and how many described established biomarkers used in grading systems as opposed to recently proposed or potentially new biomarkers. We also compared their prognostic value for late-stage wet and dry AMD against an established clinical grading system and a demographic baseline model.</p></div><div><h3>Results</h3><p>Overall, both teams independently identified clearly distinct characteristics in 27 of 30 clusters, of which 23 were related to AMD. Seven were recognized as known biomarkers used in established grading systems, and 16 depicted biomarker combinations or subtypes that are either not yet used in grading systems, were only recently proposed, or were unknown. Clusters separated incomplete from complete retinal atrophy, intraretinal from subretinal fluid, and thick from thin choroids, and, in simulation, outperformed clinically used grading systems in prognostic value.</p></div><div><h3>Conclusions</h3><p>Using self-supervised deep learning, we were able to automatically propose AMD biomarkers going beyond the set used in clinically established grading systems. Without any clinical annotations, contrastive learning discovered subtle differences between fine-grained biomarkers. Ultimately, we envision that equipping clinicians with discovery-oriented deep learning tools can accelerate the discovery of novel prognostic biomarkers.</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":"4 6","pages":"Article 100543"},"PeriodicalIF":3.2,"publicationDate":"2024-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524000794/pdfft?md5=871d81d60ba0cb5d1ab13c6991aa394a&pid=1-s2.0-S2666914524000794-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141949712","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":"Genetic Reasons for Phenotypic Diversity in Neuronal Ceroid Lipofuscinoses and High-Resolution Imaging as a Marker of Retinal Disease","authors":"Jennifer Huey LCGC ,&nbsp;Pankhuri Gupta LCGC ,&nbsp;Benjamin Wendel MS ,&nbsp;Teng Liu MS ,&nbsp;Palash Bharadwaj PhD ,&nbsp;Hillary Schwartz BS ,&nbsp;John P. Kelly PhD ,&nbsp;Irene Chang MD ,&nbsp;Jennifer R. Chao MD, PhD ,&nbsp;Ramkumar Sabesan PhD ,&nbsp;Aaron Nagiel MD, PhD ,&nbsp;Debarshi Mustafi MD, PhD","doi":"10.1016/j.xops.2024.100560","DOIUrl":"10.1016/j.xops.2024.100560","url":null,"abstract":"&lt;div&gt;&lt;h3&gt;Purpose&lt;/h3&gt;&lt;p&gt;To describe the clinical characteristics, natural history, genetic landscape, and phenotypic spectrum of neuronal ceroid lipofuscinosis (NCL)-associated retinal disease.&lt;/p&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Design&lt;/h3&gt;&lt;p&gt;Multicenter retrospective cohort study complemented by a cross-sectional examination.&lt;/p&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Subjects&lt;/h3&gt;&lt;p&gt;Twelve pediatric subjects with biallelic variants in 5 NCL-causing genes (CLN3 lysosomal/endosomal transmembrane protein [&lt;em&gt;CLN3&lt;/em&gt;], CLN6 transmembrane ER protein [&lt;em&gt;CLN6&lt;/em&gt;], Major facilitator superfamily domain containing 8 [&lt;em&gt;MFSD8&lt;/em&gt;], Palmitoyl-protein thioesterase 1 ([&lt;em&gt;PPT1&lt;/em&gt;], and tripeptidyl peptidase 1 [&lt;em&gt;TPP1&lt;/em&gt;]).&lt;/p&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Methods&lt;/h3&gt;&lt;p&gt;Review of clinical notes, retinal imaging, electroretinography (ERG), and molecular genetic testing. Two subjects underwent a cross-sectional examination comprising adaptive optics scanning laser ophthalmoscopy imaging of the retina and optoretinography (ORG).&lt;/p&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Main Outcome Measures&lt;/h3&gt;&lt;p&gt;Clinical/demographic data, multimodal retinal imaging data, electrophysiology parameters, and molecular genetic testing.&lt;/p&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Results&lt;/h3&gt;&lt;p&gt;Our cohort included a diverse set of subjects with &lt;em&gt;CLN3&lt;/em&gt;-juvenile NCL (n = 3), &lt;em&gt;TPP1&lt;/em&gt;-late infantile NCL (n = 5), &lt;em&gt;PPT1&lt;/em&gt;-late infantile or juvenile NCL (n = 2), &lt;em&gt;CLN6&lt;/em&gt;-infantile NCL (n = 1), and &lt;em&gt;CLN7&lt;/em&gt;/&lt;em&gt;MFSD8&lt;/em&gt;-late infantile NCL (n = 1). Five novel pathogenic or likely pathogenic variants were identified. Age at presentation ranged from 2 to 16 years old (mean 7.9 years). Subjects presented with varying phenotypes ranging from severe neurocognitive features (n = 8; 67%), including seizures and developmental delays and regressions, to nonsyndromic retinal dystrophies (n = 2; 17%). Visual acuities at presentation ranged from light perception to 20/20. In those with recordable ERGs, the traces were electronegative and suggestive of early cone dysfunction. Fundus imaging and OCTs demonstrated outer retinal loss that varied with underlying genotype. High-resolution adaptive optics imaging and functional measures with ORG in 2 subjects with atypical &lt;em&gt;TPP1&lt;/em&gt;-associated disease revealed significantly different phenotypes of cellular structure and function that could be followed longitudinally.&lt;/p&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Conclusions&lt;/h3&gt;&lt;p&gt;Our cohort data demonstrates that the underlying genetic variants drive the phenotypic diversity in different forms of NCL. Genetic testing can provide molecular diagnosis and ensure appropriate disease management and support for children and their families. With intravitreal enzyme replacement therapy on the horizon as a potential treatment option for NCL-associated retinal degeneration, precise structural and functional measures will be required to more accurately monitor disease progression. We show that adaptive optics imaging and ORG can be used as highly sensitive methods to track early retinal changes, whi","PeriodicalId":74363,"journal":{"name":"Ophthalmology science","volume":"4 6","pages":"Article 100560"},"PeriodicalIF":3.2,"publicationDate":"2024-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666914524000964/pdfft?md5=d02d2566e6e6217e5db3c4eb6934d69b&pid=1-s2.0-S2666914524000964-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141637213","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}