Peiyu Wang1, Jian Shen1, Ryuna Chang2, Maemae Moloney3, Mina Torres4, Bruce Burkemper2, Xuejuan Jiang2, Damien Rodger2, Rohit Varma4, Grace M Richter5. 1. Department of Biomedical Engineering, University of Southern California, Los Angeles, California. 2. USC Roski Eye Institute, Keck School of Medicine, University of Southern California, Los Angeles, California. 3. Department of Neuroscience, University of Southern California, Los Angeles, California. 4. Southern California Eyecare and Vision Research Institute, CHA Hollywood Presbyterian Medical Center, Los Angeles, California. 5. USC Roski Eye Institute, Keck School of Medicine, University of Southern California, Los Angeles, California. Electronic address: grace.richter@med.usc.edu.
Abstract
PURPOSE: To assess the diagnostic accuracy of multiple machine learning models using full retinal nerve fiber layer (RNFL) thickness maps in detecting glaucoma. DESIGN: Case-control study. PARTICIPANTS: A total of 93 eyes from 69 patients with glaucoma and 128 eyes from 128 age- and sex-matched healthy controls from the Los Angeles Latino Eye Study (LALES), a large population-based, longitudinal cohort study consisting of Latino participants aged ≥40 years residing in El Puente, California. METHODS: The 6×6-mm RNFL thickness maps centered on the optic nerve head (Cirrus 4000; Zeiss, Dublin, CA) were supplied to 4 different machine learning algorithms. These models included 2 conventional machine learning algorithms, Support Vector Machine (SVM) and K-Nearest Neighbor (KNN), and 2 convolutional neural nets, ResNet-18 and GlaucomaNet, which was a custom-made deep learning network. All models were tested with 5-fold cross validation. MAIN OUTCOME MEASURES: Area under the curve (AUC) statistics to assess diagnostic accuracy of each model compared with conventional average circumpapillary RNFL thickness. RESULTS: All 4 models achieved similarly high diagnostic accuracies, with AUC values ranging from 0.91 to 0.92. These values were significantly higher than those for average circumpapillary RNFL thickness, which had an AUC of 0.76 in the same patient population. CONCLUSIONS: Superior diagnostic performance was achieved with both conventional machine learning and convolutional neural net models compared with circumpapillary RNFL thickness. This supports the importance of the spatial structure of RNFL thickness map data in diagnosing glaucoma and further efforts to optimize our use of this data.
PURPOSE: To assess the diagnostic accuracy of multiple machine learning models using full retinal nerve fiber layer (RNFL) thickness maps in detecting glaucoma. DESIGN: Case-control study. PARTICIPANTS: A total of 93 eyes from 69 patients with glaucoma and 128 eyes from 128 age- and sex-matched healthy controls from the Los Angeles Latino Eye Study (LALES), a large population-based, longitudinal cohort study consisting of Latino participants aged ≥40 years residing in El Puente, California. METHODS: The 6×6-mm RNFL thickness maps centered on the optic nerve head (Cirrus 4000; Zeiss, Dublin, CA) were supplied to 4 different machine learning algorithms. These models included 2 conventional machine learning algorithms, Support Vector Machine (SVM) and K-Nearest Neighbor (KNN), and 2 convolutional neural nets, ResNet-18 and GlaucomaNet, which was a custom-made deep learning network. All models were tested with 5-fold cross validation. MAIN OUTCOME MEASURES: Area under the curve (AUC) statistics to assess diagnostic accuracy of each model compared with conventional average circumpapillary RNFL thickness. RESULTS: All 4 models achieved similarly high diagnostic accuracies, with AUC values ranging from 0.91 to 0.92. These values were significantly higher than those for average circumpapillary RNFL thickness, which had an AUC of 0.76 in the same patient population. CONCLUSIONS: Superior diagnostic performance was achieved with both conventional machine learning and convolutional neural net models compared with circumpapillary RNFL thickness. This supports the importance of the spatial structure of RNFL thickness map data in diagnosing glaucoma and further efforts to optimize our use of this data.
Authors: Christopher K S Leung; Shi Lam; Robert N Weinreb; Shu Liu; Cong Ye; Lan Liu; Jing He; Gilda W K Lai; Taiping Li; Dennis S C Lam Journal: Ophthalmology Date: 2010-07-21 Impact factor: 12.079
Authors: Rohit Varma; Sylvia H Paz; Stanley P Azen; Ronald Klein; Denise Globe; Mina Torres; Chrisandra Shufelt; Susan Preston-Martin Journal: Ophthalmology Date: 2004-06 Impact factor: 12.079
Authors: Hoo-Chang Shin; Holger R Roth; Mingchen Gao; Le Lu; Ziyue Xu; Isabella Nogues; Jianhua Yao; Daniel Mollura; Ronald M Summers Journal: IEEE Trans Med Imaging Date: 2016-02-11 Impact factor: 10.048
Authors: Catherine Boden; Kwokleung Chan; Pamela A Sample; Jiucang Hao; Te-Wan Lee; Linda M Zangwill; Robert N Weinreb; Michael H Goldbaum Journal: Invest Ophthalmol Vis Sci Date: 2007-12 Impact factor: 4.799
Authors: Ya Xing Wang; Hongli Yang; Haomin Luo; Seung Woo Hong; Stuart K Gardiner; Jin Wook Jeoung; Christy Hardin; Glen P Sharpe; Kouros Nouri-Mahdavi; Joseph Caprioli; Shaban Demirel; Christopher A Girkin; Jeffrey M Liebmann; Christian Y Mardin; Harry A Quigley; Alexander F Scheuerle; Brad Fortune; Balwantray C Chauhan; Claude F Burgoyne Journal: Am J Ophthalmol Date: 2020-04-13 Impact factor: 5.488