Pub Date : 2024-09-14DOI: 10.1101/2024.09.13.24313607
Maziar Mirsalehi, Benjamin Fassbind, Andreas Streich, Achim Langenbucher
Purpose Prediction of Ectasia Screening Index (ESI), an estimator provided by the Casia2 for identifying keratoconus, from raw Optical Coherence Tomography (OCT) data with Convolutional Neural Networks (CNN). Methods Three CNN architectures (ResNet18, DenseNet121 and EfficientNetB0) were employed to predict the ESI. Mean Absolute Error (MAE) was used as the performance metric for predicting the ESI by the adapted CNN models on the test set. Scans with an ESI value higher than a certain threshold were classified as Keratoconus, while the remaining scans were classified as Not Keratoconus. The models’ performance was evaluated using metrics such as accuracy, sensitivity, specificity, Positive Predictive Value (PPV) and F1 score on data collected from patients examined at the eye clinic of the Homburg University Hospital. The raw data from the Casia2 device, in 3dv format, was converted into 16 images per examination of one eye. For the training, validation and testing phases, 3689, 1050 and 1078 scans (3dv files) were selected, respectively. Results In the prediction of the ESI, the MAE values for the adapted ResNet18, DenseNet121 and EfficientNetB0, rounded to two decimal places, were 7.15, 6.64 and 5.86, respectively. In the classification task, the three networks yielded an accuracy of 94.80%, 95.27% and 95.83%, respectively; a sensitivity of 92.07%, 94.64% and 94.17%, respectively; a specificity of 96.61%, 95.69% and 96.92%, respectively; a PPV of 94.72%, 93.55% and 95.28%, respectively; and a F1 score of 93.38%, 94.09% and 94.72%, respectively. Conclusions Our results show that the prediction of keratokonus based on the ESI values estimated from raw data outperforms previous approaches using processed data. Adapted EfficientNetB0 outperformed both the other adapted models and those in state-of-the-art studies, with the highest accuracy and F1 score.
目的利用卷积神经网络（CNN）从原始光学相干断层扫描（OCT）数据中预测外生殖器筛查指数（ESI），ESI 是 Casia2 提供的用于识别角膜病的估计值。采用平均绝对误差（MAE）作为测试集上经过调整的 CNN 模型预测 ESI 的性能指标。ESI值高于一定阈值的扫描被归类为角膜病，其余扫描则被归类为非角膜病。对模型的性能评估采用了准确度、灵敏度、特异性、正预测值（PPV）和 F1 分数等指标，数据收集自在霍姆堡大学医院眼科诊所接受检查的患者。Casia2 设备提供的 3dv 格式原始数据在每次检查中被转换成 16 幅单眼图像。在训练、验证和测试阶段，分别选择了 3689、1050 和 1078 个扫描（3dv 文件）。结果在预测 ESI 时，经过调整的 ResNet18、DenseNet121 和 EfficientNetB0 的 MAE 值（四舍五入到小数点后两位）分别为 7.15、6.64 和 5.86。在分类任务中，三个网络的准确率分别为 94.80%、95.27% 和 95.83%；灵敏度分别为 92.07%、94.64% 和 94.17%；特异性分别为 96.61%、95.69% 和 96.92%；PPV 分别为 94.72%、93.结论我们的结果表明，基于原始数据估算的 ESI 值预测角膜病的效果优于之前使用处理数据的方法。改编后的 EfficientNetB0 在准确率和 F1 分数方面均优于其他改编模型和最新研究中的模型。
{"title":"Prediction of the ectasia screening index from raw Casia2 volume data for keratoconus identification by using convolutional neural networks","authors":"Maziar Mirsalehi, Benjamin Fassbind, Andreas Streich, Achim Langenbucher","doi":"10.1101/2024.09.13.24313607","DOIUrl":"https://doi.org/10.1101/2024.09.13.24313607","url":null,"abstract":"Purpose\u0000Prediction of Ectasia Screening Index (ESI), an estimator provided by the Casia2 for identifying keratoconus, from raw Optical Coherence Tomography (OCT) data with Convolutional Neural Networks (CNN).\u0000Methods\u0000Three CNN architectures (ResNet18, DenseNet121 and EfficientNetB0) were employed to predict the ESI. Mean Absolute Error (MAE) was used as the performance metric for predicting the ESI by the adapted CNN models on the test set. Scans with an ESI value higher than a certain threshold were classified as Keratoconus, while the remaining scans were classified as Not Keratoconus. The models’ performance was evaluated using metrics such as accuracy, sensitivity, specificity, Positive Predictive Value (PPV) and F1 score on data collected from patients examined at the eye clinic of the Homburg University Hospital. The raw data from the Casia2 device, in 3dv format, was converted into 16 images per examination of one eye. For the training, validation and testing phases, 3689, 1050 and 1078 scans (3dv files) were selected, respectively. Results\u0000In the prediction of the ESI, the MAE values for the adapted ResNet18, DenseNet121 and EfficientNetB0, rounded to two decimal places, were 7.15, 6.64 and 5.86, respectively. In the classification task, the three networks yielded an accuracy of 94.80%, 95.27% and 95.83%, respectively; a sensitivity of 92.07%, 94.64% and 94.17%, respectively; a specificity of 96.61%, 95.69% and 96.92%, respectively; a PPV of 94.72%, 93.55% and 95.28%, respectively; and a F1 score of 93.38%, 94.09% and 94.72%, respectively.\u0000Conclusions\u0000Our results show that the prediction of keratokonus based on the ESI values estimated from raw data outperforms previous approaches using processed data. Adapted EfficientNetB0 outperformed both the other adapted models and those in state-of-the-art studies, with the highest accuracy and F1 score.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260260","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}
Pub Date : 2024-09-13DOI: 10.1101/2024.09.12.24313551
Prashant D. Tailor, Haley S. D'Souza, Clara Castillejo Becerra, Heidi M. Dahl, Neil R. Patel, Tyler M. Kaplan, Darrell Kohli, Erick D. Bothun, Brian G. Mohney, Andrea A. Tooley, Keith H. Baratz, Raymond Iezzi, Andrew J. Barkmeier, Sophie J. Bakri, Gavin W. Roddy, David Hodge, Arthur J. Sit, Matthew R. Starr, John J. Chen
Background Specialized terminology employed by ophthalmologists creates a comprehension barrier for non-ophthalmology providers, compromising interdisciplinary communication and patient care. Current solutions such as manual note simplification are impractical or inadequate. Large language models (LLMs) present a potential low-burden approach to translating ophthalmology documentation into accessible language. Methods This prospective, randomized trial evaluated the addition of LLM-generated plain language summaries (PLSs) to standard ophthalmology notes (SONs). Participants included non-ophthalmology providers and ophthalmologists. The study assessed: (1) non-ophthalmology providers' comprehension and satisfaction with either the SON (control) or SON+PLS (intervention), (2) ophthalmologists' evaluation of PLS accuracy, safety, and time burden, and (3) objective semantic and linguistic quality of PLSs. Results 85% of non-ophthalmology providers (n=362, 33% response rate) preferred the PLS to SON. Non-ophthalmology providers reported enhanced diagnostic understanding (p=0.012), increased note detail satisfaction (p<0.001), and improved explanation clarity (p<0.001) for notes containing a PLS. The addition of a PLS narrowed comprehension gaps between providers who were comfortable and uncomfortable with ophthalmology terminology at baseline (intergroup difference p<0.001 to p>0.05). PLS semantic analysis demonstrated high meaning preservation (BERTScore mean F1 score: 0.85) with greater readability (Flesch Reading Ease: 51.8 vs. 43.6, Flesch-Kincaid Grade Level: 10.7 vs. 11.9). Ophthalmologists (n=489, 84% response rate) reported high PLS accuracy (90% "a great deal") with minimal review time burden (94.9% ≤ 1 minute). PLS error rate on initial ophthalmologist review and editing was 26%, and 15% on independent ophthalmologist over-read of edited PLSs. 84.9% of identified errors were deemed low risk for patient harm and 0% had a risk of severe harm/death. Conclusions LLM-generated plain language summaries enhance accessibility and utility of ophthalmology notes for non-ophthalmology providers while maintaining high semantic fidelity and improving readability. PLS error rates underscore the need for careful implementation and ongoing safety monitoring in clinical practice.
{"title":"Utilizing AI-Generated Plain Language Summaries to Enhance Interdisciplinary Understanding of Ophthalmology Notes: A Randomized Trial","authors":"Prashant D. Tailor, Haley S. D'Souza, Clara Castillejo Becerra, Heidi M. Dahl, Neil R. Patel, Tyler M. Kaplan, Darrell Kohli, Erick D. Bothun, Brian G. Mohney, Andrea A. Tooley, Keith H. Baratz, Raymond Iezzi, Andrew J. Barkmeier, Sophie J. Bakri, Gavin W. Roddy, David Hodge, Arthur J. Sit, Matthew R. Starr, John J. Chen","doi":"10.1101/2024.09.12.24313551","DOIUrl":"https://doi.org/10.1101/2024.09.12.24313551","url":null,"abstract":"Background Specialized terminology employed by ophthalmologists creates a comprehension barrier for non-ophthalmology providers, compromising interdisciplinary communication and patient care. Current solutions such as manual note simplification are impractical or inadequate. Large language models (LLMs) present a potential low-burden approach to translating ophthalmology documentation into accessible language. Methods This prospective, randomized trial evaluated the addition of LLM-generated plain language summaries (PLSs) to standard ophthalmology notes (SONs). Participants included non-ophthalmology providers and ophthalmologists. The study assessed: (1) non-ophthalmology providers' comprehension and satisfaction with either the SON (control) or SON+PLS (intervention), (2) ophthalmologists' evaluation of PLS accuracy, safety, and time burden, and (3) objective semantic and linguistic quality of PLSs. Results 85% of non-ophthalmology providers (n=362, 33% response rate) preferred the PLS to SON. Non-ophthalmology providers reported enhanced diagnostic understanding (p=0.012), increased note detail satisfaction (p<0.001), and improved explanation clarity (p<0.001) for notes containing a PLS. The addition of a PLS narrowed comprehension gaps between providers who were comfortable and uncomfortable with ophthalmology terminology at baseline (intergroup difference p<0.001 to p>0.05). PLS semantic analysis demonstrated high meaning preservation (BERTScore mean F1 score: 0.85) with greater readability (Flesch Reading Ease: 51.8 vs. 43.6, Flesch-Kincaid Grade Level: 10.7 vs. 11.9). Ophthalmologists (n=489, 84% response rate) reported high PLS accuracy (90% \"a great deal\") with minimal review time burden (94.9% ≤ 1 minute). PLS error rate on initial ophthalmologist review and editing was 26%, and 15% on independent ophthalmologist over-read of edited PLSs. 84.9% of identified errors were deemed low risk for patient harm and 0% had a risk of severe harm/death. Conclusions LLM-generated plain language summaries enhance accessibility and utility of ophthalmology notes for non-ophthalmology providers while maintaining high semantic fidelity and improving readability. PLS error rates underscore the need for careful implementation and ongoing safety monitoring in clinical practice.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260037","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}
Pub Date : 2024-09-12DOI: 10.1101/2024.09.11.24313484
Sarika Gopalakrishnan, Ronald Schuchard, Donald Fletcher, Arathy Kartha
We compared visual function in individuals with low vision (>20/60) using three different models of augmented reality (AR) devices: Ziru, IrisVision, and NuEyes-Pro3. The distance visual acuity (VA) was measured in high luminance high contrast (HLHC), high luminance low contrast (HLLC), low luminance high contrast (LLHC), and low luminance low contrast (LLLC) settings. The other tests were near VA, distance and near contrast sensitivity (CS), color vision, depth perception and indoor navigation. The change in visual function without and with AR devices was analyzed. Out of 27 participants, 17 were female. The mean age was 66.7 (18.2) years. The median baseline VA was 0.66 (0.49) logMAR in HLHC, 0.87 (0.54) logMAR in HLLC, 0.84 (0.67) logMAR in LLHC and 1.04 (0.34) logMAR in LLLC. The median baseline near VA was 0.55(0.4) logMAR, distance and near CS was 1.10(0.26) logCS, and 1.20(0.30) logCS respectively. Distance and near vision showed significant differences with both Ziru and IrisVision (p<0.01), but not with NuEyes. There was a significant change in CS using Ziru and IrisVision for both distance and near (p<0.05) but both reduced significantly with NuEyes (p<0.01). The baseline functional vision score (FVS) was 45.76 (44.47) which improved to 79.04 (33.98) with Ziru and 76.14 (33.76) with IrisVision significantly, whereas it significantly reduced to 35.00 (33.97) with NuEyes (p<0.01). During the objective identification task on the indoor mobility course using AR devices, head-level objects were missed more compared to waist or floor-level objects across all three models. Majority of the visual functions improved with Ziru and IrisVision, with limited improvement in certain lighting condition of distance visual acuity with NuEyes.
{"title":"Comparison of visual function analysis of people with low vision using three different models of augmented reality devices","authors":"Sarika Gopalakrishnan, Ronald Schuchard, Donald Fletcher, Arathy Kartha","doi":"10.1101/2024.09.11.24313484","DOIUrl":"https://doi.org/10.1101/2024.09.11.24313484","url":null,"abstract":"We compared visual function in individuals with low vision (>20/60) using three different models of augmented reality (AR) devices: Ziru, IrisVision, and NuEyes-Pro3. The distance visual acuity (VA) was measured in high luminance high contrast (HLHC), high luminance low contrast (HLLC), low luminance high contrast (LLHC), and low luminance low contrast (LLLC) settings. The other tests were near VA, distance and near contrast sensitivity (CS), color vision, depth perception and indoor navigation. The change in visual function without and with AR devices was analyzed. Out of 27 participants, 17 were female. The mean age was 66.7 (18.2) years. The median baseline VA was 0.66 (0.49) logMAR in HLHC, 0.87 (0.54) logMAR in HLLC, 0.84 (0.67) logMAR in LLHC and 1.04 (0.34) logMAR in LLLC. The median baseline near VA was 0.55(0.4) logMAR, distance and near CS was 1.10(0.26) logCS, and 1.20(0.30) logCS respectively. Distance and near vision showed significant differences with both Ziru and IrisVision (p<0.01), but not with NuEyes. There was a significant change in CS using Ziru and IrisVision for both distance and near (p<0.05) but both reduced significantly with NuEyes (p<0.01). The baseline functional vision score (FVS) was 45.76 (44.47) which improved to 79.04 (33.98) with Ziru and 76.14 (33.76) with IrisVision significantly, whereas it significantly reduced to 35.00 (33.97) with NuEyes (p<0.01). During the objective identification task on the indoor mobility course using AR devices, head-level objects were missed more compared to waist or floor-level objects across all three models. Majority of the visual functions improved with Ziru and IrisVision, with limited improvement in certain lighting condition of distance visual acuity with NuEyes.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217443","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}
Pub Date : 2024-09-12DOI: 10.1101/2024.09.11.24312817
Himeesh Kumar, Yelena Bagdasarova, Scott Song, Doron G. Hickey, Amy C. Cohn, Mali Okada, Robert P. Finger, Jan H. Terheyden, Ruth E. Hogg, Pierre-Henry Gabrielle, Louis Arnould, Maxime Jannaud, Xavier Hadoux, Peter van Wijngaarden, Carla J. Abbott, Lauren A.B. Hodgson, Roy Schwartz, Adnan Tufail, Emily Y. Chew, Cecilia S. Lee, Erica L. Fletcher, Melanie Bahlo, Brendan R.E. Ansell, Alice Pebay, Robyn H. Guymer, Aaron Y. Lee, Zhichao Wu
Reticular pseudodrusen (RPD) signify a critical phenotype driving vision loss in age-related macular degeneration (AMD). Their detection is paramount in the clinical management of those with AMD, yet they remain challenging to reliably identify. We thus developed a deep learning (DL) model to segment RPD from 9,800 optical coherence tomography B-scans, and this model produced RPD segmentations that had higher agreement with four retinal specialists (Dice similarity coefficient [DSC]=0.76 [95% confidence interval [CI] 0.71-0.81]) than the agreement amongst the specialists (DSC=0.68, 95% CI=0.63-0.73; p<0.001). In five external test datasets consisting of 1,017 eyes from 812 individuals, the DL model detected RPD with a similar level of performance as two retinal specialists (area-under-the-curve of 0.94 [95% CI=0.92-0.97], 0.95 [95% CI=0.92-0.97] and 0.96 [95% CI=0.94-0.98] respectively; p≥0.32). This DL model enables the automatic detection and quantification of RPD with expert-level performance, which we have made publicly available.
{"title":"Deep Learning-Based Detection of Reticular Pseudodrusen in Age-Related Macular Degeneration on Optical Coherence Tomography","authors":"Himeesh Kumar, Yelena Bagdasarova, Scott Song, Doron G. Hickey, Amy C. Cohn, Mali Okada, Robert P. Finger, Jan H. Terheyden, Ruth E. Hogg, Pierre-Henry Gabrielle, Louis Arnould, Maxime Jannaud, Xavier Hadoux, Peter van Wijngaarden, Carla J. Abbott, Lauren A.B. Hodgson, Roy Schwartz, Adnan Tufail, Emily Y. Chew, Cecilia S. Lee, Erica L. Fletcher, Melanie Bahlo, Brendan R.E. Ansell, Alice Pebay, Robyn H. Guymer, Aaron Y. Lee, Zhichao Wu","doi":"10.1101/2024.09.11.24312817","DOIUrl":"https://doi.org/10.1101/2024.09.11.24312817","url":null,"abstract":"Reticular pseudodrusen (RPD) signify a critical phenotype driving vision loss in age-related macular degeneration (AMD). Their detection is paramount in the clinical management of those with AMD, yet they remain challenging to reliably identify. We thus developed a deep learning (DL) model to segment RPD from 9,800 optical coherence tomography B-scans, and this model produced RPD segmentations that had higher agreement with four retinal specialists (Dice similarity coefficient [DSC]=0.76 [95% confidence interval [CI] 0.71-0.81]) than the agreement amongst the specialists (DSC=0.68, 95% CI=0.63-0.73; p<0.001). In five external test datasets consisting of 1,017 eyes from 812 individuals, the DL model detected RPD with a similar level of performance as two retinal specialists (area-under-the-curve of 0.94 [95% CI=0.92-0.97], 0.95 [95% CI=0.92-0.97] and 0.96 [95% CI=0.94-0.98] respectively; p≥0.32). This DL model enables the automatic detection and quantification of RPD with expert-level performance, which we have made publicly available.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217423","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}
Pub Date : 2024-09-12DOI: 10.1101/2024.09.10.24313265
Anupam K. Garg, Jingyu Wang, Bailee Alonzo, Ji Yi, Amir H. Kashani
Purpose: To evaluate outer retinal organization in normal subjects and those using hydroxychloroquine (HCQ) with ultrahigh resolution visible light optical coherence tomography (VIS-OCT). Methods: Forty eyes of 22 adult subjects were recruited from a tertiary care retina practice including controls (20 eyes, 12 subjects, mean age 40±22yrs, mean logMAR BCVA 0.19, 90% female) and subjects with a history of HCQ use (20 eyes, 10 subjects, mean age 62±17yrs, mean logMAR BCVA 0.03, 67% female). Each subject was imaged using a custom-built VIS-OCT device (axial resolution 1.3μm) and FDA-approved OCT devices. Results: Using VIS-OCT, control subjects demonstrate 5 and 6 hyperreflective bands in the foveal and parafoveal regions, respectively, between the outer nuclear layer and Bruch ′s membrane. These bands demonstrate intensity profiles complementary to the known histopathologic distribution of rods and cones. In comparison to controls, subjects taking HCQ demonstrate blunting of all bands, particularly bands 2-4. In all cases of suspected or known toxicity, VIS-OCT demonstrated attenuation of band 3i and in no cases was there attenuation of other bands that was more severe than band 3i, suggesting that changes in the reflectivity of Band 3i may be the earliest identifiable sign of HCQ toxicity. Conclusions: VIS-OCT of the outer retina demonstrates a unique outer retinal banding pattern corresponding to photoreceptor density profiles. There is a notable attenuation of the photoreceptor outer segment reflectivity profile associated with early HCQ toxicity. This finding may be an early, and possibly reversible, sign of HCQ toxicity, primarily impacting the cones.
{"title":"Photoreceptor outer segment reflectivity with ultrahigh resolution visible light optical coherence tomography in systemic hydroxychloroquine use","authors":"Anupam K. Garg, Jingyu Wang, Bailee Alonzo, Ji Yi, Amir H. Kashani","doi":"10.1101/2024.09.10.24313265","DOIUrl":"https://doi.org/10.1101/2024.09.10.24313265","url":null,"abstract":"Purpose: To evaluate outer retinal organization in normal subjects and those using hydroxychloroquine (HCQ) with ultrahigh resolution visible light optical coherence tomography (VIS-OCT).\u0000Methods: Forty eyes of 22 adult subjects were recruited from a tertiary care retina practice including controls (20 eyes, 12 subjects, mean age 40±22yrs, mean logMAR BCVA 0.19, 90% female) and subjects with a history of HCQ use (20 eyes, 10 subjects, mean age 62±17yrs, mean logMAR BCVA 0.03, 67% female). Each subject was imaged using a custom-built VIS-OCT device (axial resolution 1.3μm) and FDA-approved OCT devices.\u0000Results: Using VIS-OCT, control subjects demonstrate 5 and 6 hyperreflective bands in the foveal and parafoveal regions, respectively, between the outer nuclear layer and Bruch ′s membrane. These bands demonstrate intensity profiles complementary to the known histopathologic distribution of rods and cones. In comparison to controls, subjects taking HCQ demonstrate blunting of all bands, particularly bands 2-4. In all cases of suspected or known toxicity, VIS-OCT demonstrated attenuation of band 3i and in no cases was there attenuation of other bands that was more severe than band 3i, suggesting that changes in the reflectivity of Band 3i may be the earliest identifiable sign of HCQ toxicity.\u0000Conclusions: VIS-OCT of the outer retina demonstrates a unique outer retinal banding pattern corresponding to photoreceptor density profiles. There is a notable attenuation of the photoreceptor outer segment reflectivity profile associated with early HCQ toxicity. This finding may be an early, and possibly reversible, sign of HCQ toxicity, primarily impacting the cones.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217442","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}
Pub Date : 2024-09-10DOI: 10.1101/2024.09.09.24313319
Yuan-Kai Fu, Matthew Lin, Kuo-Hsuan Hung, Lung-Kun Yeh, HsinYuan Tan
Corneal endothelial dysfunction is the main cause for more than 50% of corneal transplantations. Human corneal endothelial cells are generally viewed as non-proliferative in vivo. Any injury that results in endothelial loss exceeding the critical threshold can cause irreversible endothelial functional decompensation, leading to corneal edema and vision loss. Currently, the mainstay treatment for irreversible corneal dysfunction is corneal transplantation. In this work, using well-established imaging technique of specular microscopy, we revisited the endothelial damage following three common corneal endothelial injury scenarios: post-cataract surgery, endothelial dystrophy, and corneal penetrating injury. We identified unexpected, stochastic single-cell loss in the corneal endothelium following primary injuries, persisting well beyond the expected wound healing period, a phenomenon that has not been previously highlighted. This finding offers a potential explanation for the chronic endothelial cell loss following a primary injury. Further investigation could provide valuable insights for improving clinical management strategies for corneal endothelial dysfunction.
{"title":"Excessive Corneal Endothelial Single Cell Loss Following Endothelial Injuries","authors":"Yuan-Kai Fu, Matthew Lin, Kuo-Hsuan Hung, Lung-Kun Yeh, HsinYuan Tan","doi":"10.1101/2024.09.09.24313319","DOIUrl":"https://doi.org/10.1101/2024.09.09.24313319","url":null,"abstract":"Corneal endothelial dysfunction is the main cause for more than 50% of corneal transplantations. Human corneal endothelial cells are generally viewed as non-proliferative in vivo. Any injury that results in endothelial loss exceeding the critical threshold can cause irreversible endothelial functional decompensation, leading to corneal edema and vision loss. Currently, the mainstay treatment for irreversible corneal dysfunction is corneal transplantation. In this work, using well-established imaging technique of specular microscopy, we revisited the endothelial damage following three common corneal endothelial injury scenarios: post-cataract surgery, endothelial dystrophy, and corneal penetrating injury. We identified unexpected, stochastic single-cell loss in the corneal endothelium following primary injuries, persisting well beyond the expected wound healing period, a phenomenon that has not been previously highlighted. This finding offers a potential explanation for the chronic endothelial cell loss following a primary injury. Further investigation could provide valuable insights for improving clinical management strategies for corneal endothelial dysfunction.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217424","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}
Objective This study investigates the potential of blood biomarkers in the early diagnosis of glaucoma, focusing on erythrocyte osmotic fragility (EOF) as a novel indicator. We used Least Absolute Shrinkage and Selection Operator (LASSO) regression to evaluate EOF’s predictive utility for glaucoma diagnosis.
{"title":"Erythrocyte Osmotic Fragility as A Diagnostic Marker in Glaucoma: A Comprehensive Analysis Using LASSO Regression","authors":"Jialiang Yang, Fang Yang, Junming Gu, Yilian Cheng, Qian Luo, Fang Hao, Bo Gong, Houbin Zhang","doi":"10.1101/2024.09.05.24312988","DOIUrl":"https://doi.org/10.1101/2024.09.05.24312988","url":null,"abstract":"<strong>Objective</strong> This study investigates the potential of blood biomarkers in the early diagnosis of glaucoma, focusing on erythrocyte osmotic fragility (EOF) as a novel indicator. We used Least Absolute Shrinkage and Selection Operator (LASSO) regression to evaluate EOF’s predictive utility for glaucoma diagnosis.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217425","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}
This comprehensive study investigates the gender distribution of ophthalmologists in academic positions in Japanese medical schools. Data were collected from the websites of all ophthalmology departments and affiliated hospitals from November 1-15, 2023. Faculty gender was identified using first names and photographs, and where unclear, further verification was sought from the Ministry of Health, Labour, and Welfare or known contacts. The survey included 1,574 faculty members: 453 females (28.8%) and 1,121 males (71.2%). The representation of females varied significantly across different academic ranks: 9.9% were professors, 21.9% associate professors, 39.5% assistant professors, and 36.7% research associates. A significant gender disparity was observed, with males being 4.41 times more likely to become professors than females (p < 0.001). Conversely, females were more likely to hold research associate positions than males (Odds Ratio: 0.51, p < 0.001). This study highlights a male predominance in senior academic positions within ophthalmology departments in Japanese medical schools.
{"title":"Investigating Gender Disparities in Ophthalmology Departments at Medical Schools in Japan","authors":"Akemi Iwasaki, Naoko Kato, Yuka Morita, Hiromi Onouchi, Mariko Itakura, Keiko Kunimi, Yoichi Manabe","doi":"10.1101/2024.09.01.24312900","DOIUrl":"https://doi.org/10.1101/2024.09.01.24312900","url":null,"abstract":"This comprehensive study investigates the gender distribution of ophthalmologists in academic positions in Japanese medical schools. Data were collected from the websites of all ophthalmology departments and affiliated hospitals from November 1-15, 2023. Faculty gender was identified using first names and photographs, and where unclear, further verification was sought from the Ministry of Health, Labour, and Welfare or known contacts. The survey included 1,574 faculty members: 453 females (28.8%) and 1,121 males (71.2%). The representation of females varied significantly across different academic ranks: 9.9% were professors, 21.9% associate professors, 39.5% assistant professors, and 36.7% research associates. A significant gender disparity was observed, with males being 4.41 times more likely to become professors than females (p < 0.001). Conversely, females were more likely to hold research associate positions than males (Odds Ratio: 0.51, p < 0.001). This study highlights a male predominance in senior academic positions within ophthalmology departments in Japanese medical schools.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217426","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}
Pub Date : 2024-08-28DOI: 10.1101/2024.08.26.24311677
Dalia Camacho-García-Formentí, Gabriela Baylón-Vázquez, Karen Janeth Arriozola-Rodríguez, Luis Enrique Avalos-Ramirez, Curt Hartleben-Matkin, Hugo Francisco Valdez Flores, Damaris Hodelin-Fuentes, Alejandro Noriega Campero
Background: Artificial intelligence (AI) shows promise in ophthalmology, but its performance in diverse healthcare settings remains understudied. We evaluated retinIA, an AI-powered screening tool developed with Mexican data, against first-year ophthalmology residents in a tertiary care setting in Mexico City. Methods: We analyzed 435 adult patients undergoing their first ophthalmic evaluation. RetinIA and residents' assessments were compared against expert annotations for retinal lesions, cup-to-disk ratio (CDR) measurements, and glaucoma suspect detection. We also evaluated a synergistic approach combining AI and resident assessments. Results: For glaucoma suspect detection, retinIA outperformed residents in accuracy (88.6% vs 82.9%, p = 0.016), sensitivity (63.0% vs 50.0%, p = 0.116), and specificity (94.5% vs 90.5%, p = 0.062). While, the synergistic approach deemed a higher sensitivity (80.4%) than ophthalmic residents alone or retinIA alone (p < 0.001). RetinIA's CDR estimates showed lower mean absolute error (0.056 vs 0.105, p < 0.001) and higher correlation with expert measurements (r = 0.728 vs r = 0.538). In retinal lesion detection, retinIA demonstrated superior sensitivity (90.1% vs 63.0% for medium/high-risk lesions, p < 0.001) and specificity (95.8% vs 90.4%, p < 0.001). Furthermore, differences between retinIA and residents were statistically significant across all metrics. The synergistic approach achieved the highest sensitivity for retinal lesions (92.6% for medium/high-risk, 100% for high-risk) while maintaining good specificity (87.4%). Conclusion: RetinIA outperforms first-year residents in key ophthalmic assessments. The synergistic use of AI and resident assessments shows potential for optimizing diagnostic accuracy, highlighting the value of AI as a supportive tool in ophthalmic practice, especially for early-career clinicians.
背景：人工智能（AI）在眼科领域大有可为，但其在不同医疗环境中的表现仍未得到充分研究。我们在墨西哥城的一家三级医疗机构，针对一年级眼科住院医师评估了利用墨西哥数据开发的人工智能筛查工具 retinIA：我们对首次接受眼科评估的 435 名成年患者进行了分析。我们将 RetinIA 和住院医生的评估结果与专家对视网膜病变、杯盘比（CDR）测量和青光眼疑似病例检测的注释进行了比较。我们还评估了一种结合人工智能和住院医师评估的协同方法：结果：对于青光眼疑似病例的检测，retinIA 在准确性（88.6% vs 82.9%，p = 0.016）、灵敏度（63.0% vs 50.0%，p = 0.116）和特异性（94.5% vs 90.5%，p = 0.062）方面均优于住院医生。而协同方法的灵敏度（80.4%）比单独使用眼科住院医师或单独使用视网膜内皮素IA更高（p <0.001）。RetinIA 的 CDR 估计值显示出更低的平均绝对误差（0.056 vs 0.105，p <0.001）和更高的与专家测量值的相关性（r = 0.728 vs r = 0.538）。在视网膜病变检测方面，retinIA 的灵敏度（90.1% 对 63.0%，p <0.001）和特异性（95.8% 对 90.4%，p <0.001）更胜一筹。此外，在所有指标上，视网膜内窥镜和住院医师之间的差异都具有统计学意义。协同方法对视网膜病变的敏感性最高（中/高风险为 92.6%，高风险为 100%），同时保持良好的特异性（87.4%）：RetinIA在关键眼科评估方面的表现优于一年级住院医师。人工智能和住院医师评估的协同使用显示了优化诊断准确性的潜力，凸显了人工智能作为眼科实践中的辅助工具的价值，尤其是对于初入职场的临床医师而言。
{"title":"Comparative Performance of retinIA, an AI-powered Ophthalmic Screening Tool, and First-Year Residents in Retinal Disease Detection and Glaucoma Assessment: A Study in a Mexican Tertiary Care Setting","authors":"Dalia Camacho-García-Formentí, Gabriela Baylón-Vázquez, Karen Janeth Arriozola-Rodríguez, Luis Enrique Avalos-Ramirez, Curt Hartleben-Matkin, Hugo Francisco Valdez Flores, Damaris Hodelin-Fuentes, Alejandro Noriega Campero","doi":"10.1101/2024.08.26.24311677","DOIUrl":"https://doi.org/10.1101/2024.08.26.24311677","url":null,"abstract":"Background: Artificial intelligence (AI) shows promise in ophthalmology, but its performance in diverse healthcare settings remains understudied. We evaluated retinIA, an AI-powered screening tool developed with Mexican data, against first-year ophthalmology residents in a tertiary\u0000care setting in Mexico City.\u0000Methods: We analyzed 435 adult patients undergoing their first ophthalmic evaluation. RetinIA and residents' assessments were compared\u0000against expert annotations for retinal lesions, cup-to-disk ratio (CDR) measurements, and glaucoma suspect detection. We also evaluated a synergistic approach combining AI and resident assessments.\u0000Results: For glaucoma suspect detection, retinIA outperformed residents in accuracy (88.6% vs 82.9%, p = 0.016), sensitivity (63.0% vs\u000050.0%, p = 0.116), and specificity (94.5% vs 90.5%, p = 0.062). While, the synergistic approach deemed a higher sensitivity (80.4%) than ophthalmic residents alone or retinIA alone (p < 0.001). RetinIA's CDR estimates showed lower mean absolute error (0.056 vs 0.105, p < 0.001) and\u0000higher correlation with expert measurements (r = 0.728 vs r = 0.538). In retinal lesion detection, retinIA demonstrated superior sensitivity (90.1%\u0000vs 63.0% for medium/high-risk lesions, p < 0.001) and specificity (95.8% vs 90.4%, p < 0.001). Furthermore, differences between retinIA and residents were statistically significant across all metrics. The synergistic\u0000approach achieved the highest sensitivity for retinal lesions (92.6% for medium/high-risk, 100% for high-risk) while maintaining good specificity\u0000(87.4%).\u0000Conclusion: RetinIA outperforms first-year residents in key ophthalmic assessments. The synergistic use of AI and resident assessments\u0000shows potential for optimizing diagnostic accuracy, highlighting the value\u0000of AI as a supportive tool in ophthalmic practice, especially for early-career clinicians.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217428","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}
Pub Date : 2024-08-28DOI: 10.1101/2024.08.27.24312649
Jamie Burke, Samuel Gibbon, Audrey Low, Charlene Hamid, Megan Reid-Schachter, Graciela Muniz Terrera, Craig Ritchie, Baljean Dhillon, John T O'Brien, Stuart King, Ian J.C. MacCormick, Tom J. MacGillivray
Objective: To explore associations between measurements of the ocular microvasculature in the choroid (a highly vascularised layer posterior to the retina) and genetic Alzheimer's disease risk. Methods: We measured the choroidal vasculature appearing in optical coherence tomography scans of 69 healthy, mid-life individuals in the PREVENT cohort. The cohort was prospectively split into low, medium, and high-risk groups based on the presence of known risk factors (APOE4 genotype and family history of dementia). We used ordinal logistic regression to test for cross-sectional associations between choroidal measurements and pre-determined risk of future Alzheimer's disease. Results: We observed progressively increased choroidal vasculature between ordinal risk groups, and all choroidal measurements were significantly associated with risk group prediction. APOE4 carriers had significantly thicker choroids and larger vascular tissue compared to non-carriers. Similar trends were observed for those with a family history of dementia. In our sample, a 0.16 mm2 increase in choroidal vascular area was associated with a 2-fold increase in the likelihood of having one or more markers of Alzheimer's disease risk, compared with none. Conclusions: Our results suggest a potential link between the choroidal vasculature and risk of Alzheimer's disease. However, these findings are exploratory and should be replicated in a larger, more diverse sample.
{"title":"Association between choroidal microvasculature in the eye and Alzheimer's disease risk in cognitively healthy midlife adults","authors":"Jamie Burke, Samuel Gibbon, Audrey Low, Charlene Hamid, Megan Reid-Schachter, Graciela Muniz Terrera, Craig Ritchie, Baljean Dhillon, John T O'Brien, Stuart King, Ian J.C. MacCormick, Tom J. MacGillivray","doi":"10.1101/2024.08.27.24312649","DOIUrl":"https://doi.org/10.1101/2024.08.27.24312649","url":null,"abstract":"Objective: To explore associations between measurements of the ocular microvasculature in the choroid (a highly vascularised layer posterior to the retina) and genetic Alzheimer's disease risk. Methods: We measured the choroidal vasculature appearing in optical coherence tomography scans of 69 healthy, mid-life individuals in the PREVENT cohort. The cohort was prospectively split into low, medium, and high-risk groups based on the presence of known risk factors (APOE4 genotype and family history of dementia). We used ordinal logistic regression to test for cross-sectional associations between choroidal measurements and pre-determined risk of future Alzheimer's disease. Results: We observed progressively increased choroidal vasculature between ordinal risk groups, and all choroidal measurements were significantly associated with risk group prediction. APOE4 carriers had significantly thicker choroids and larger vascular tissue compared to non-carriers. Similar trends were observed for those with a family history of dementia. In our sample, a 0.16 mm2 increase in choroidal vascular area was associated with a 2-fold increase in the likelihood of having one or more markers of Alzheimer's disease risk, compared with none. Conclusions: Our results suggest a potential link between the choroidal vasculature and risk of Alzheimer's disease. However, these findings are exploratory and should be replicated in a larger, more diverse sample.","PeriodicalId":501390,"journal":{"name":"medRxiv - Ophthalmology","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142217427","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：