Benjamin Y Xu1, Michael Chiang2, Shreyasi Chaudhary3, Shraddha Kulkarni3, Anmol A Pardeshi4, Rohit Varma5. 1. Roski Eye Institute, Department of Ophthalmology, Keck School of Medicine at the University of Southern California, Los Angeles, California, USA. Electronic address: benjamin.xu@med.usc.edu. 2. Sol Price School of Public Policy, University of Southern California, Los Angeles, California, USA. 3. Viterbi School of Engineering, University of Southern California, Los Angeles, California, USA. 4. Roski Eye Institute, Department of Ophthalmology, Keck School of Medicine at the University of Southern California, Los Angeles, California, USA. 5. Southern California Eye Institute, CHA Hollywood Presbyterian Medical Center, Los Angeles, California, USA.
Abstract
PURPOSE: To develop and test deep learning classifiers that detect gonioscopic angle closure and primary angle closure disease (PACD) based on fully automated analysis of anterior segment OCT (AS-OCT) images. METHODS: Subjects were recruited as part of the Chinese-American Eye Study (CHES), a population-based study of Chinese Americans in Los Angeles, California, USA. Each subject underwent a complete ocular examination including gonioscopy and AS-OCT imaging in each quadrant of the anterior chamber angle (ACA). Deep learning methods were used to develop 3 competing multi-class convolutional neural network (CNN) classifiers for modified Shaffer grades 0, 1, 2, 3, and 4. Binary probabilities for closed (grades 0 and 1) and open (grades 2, 3, and 4) angles were calculated by summing over the corresponding grades. Classifier performance was evaluated by 5-fold cross-validation and on an independent test dataset. Outcome measures included area under the receiver operating characteristic curve (AUC) for detecting gonioscopic angle closure and PACD, defined as either 2 or 3 quadrants of gonioscopic angle closure per eye. RESULTS: A total of 4036 AS-OCT images with corresponding gonioscopy grades (1943 open, 2093 closed) were obtained from 791 CHES subjects. Three competing CNN classifiers were developed with a cross-validation dataset of 3396 images (1632 open, 1764 closed) from 664 subjects. The remaining 640 images (311 open, 329 closed) from 127 subjects were segregated into a test dataset. The best-performing classifier was developed by applying transfer learning to the ResNet-18 architecture. For detecting gonioscopic angle closure, this classifier achieved an AUC of 0.933 (95% confidence interval, 0.925-0.941) on the cross-validation dataset and 0.928 on the test dataset. For detecting PACD based on 2- and 3-quadrant definitions, the ResNet-18 classifier achieved AUCs of 0.964 and 0.952, respectively, on the test dataset. CONCLUSION: Deep learning classifiers effectively detect gonioscopic angle closure and PACD based on automated analysis of AS-OCT images. These methods could be used to automate clinical evaluations of the ACA and improve access to eye care in high-risk populations.
PURPOSE: To develop and test deep learning classifiers that detect gonioscopic angle closure and primary angle closure disease (PACD) based on fully automated analysis of anterior segmentOCT (AS-OCT) images. METHODS: Subjects were recruited as part of the Chinese-American Eye Study (CHES), a population-based study of Chinese Americans in Los Angeles, California, USA. Each subject underwent a complete ocular examination including gonioscopy and AS-OCT imaging in each quadrant of the anterior chamber angle (ACA). Deep learning methods were used to develop 3 competing multi-class convolutional neural network (CNN) classifiers for modified Shaffer grades 0, 1, 2, 3, and 4. Binary probabilities for closed (grades 0 and 1) and open (grades 2, 3, and 4) angles were calculated by summing over the corresponding grades. Classifier performance was evaluated by 5-fold cross-validation and on an independent test dataset. Outcome measures included area under the receiver operating characteristic curve (AUC) for detecting gonioscopic angle closure and PACD, defined as either 2 or 3 quadrants of gonioscopic angle closure per eye. RESULTS: A total of 4036 AS-OCT images with corresponding gonioscopy grades (1943 open, 2093 closed) were obtained from 791 CHES subjects. Three competing CNN classifiers were developed with a cross-validation dataset of 3396 images (1632 open, 1764 closed) from 664 subjects. The remaining 640 images (311 open, 329 closed) from 127 subjects were segregated into a test dataset. The best-performing classifier was developed by applying transfer learning to the ResNet-18 architecture. For detecting gonioscopic angle closure, this classifier achieved an AUC of 0.933 (95% confidence interval, 0.925-0.941) on the cross-validation dataset and 0.928 on the test dataset. For detecting PACD based on 2- and 3-quadrant definitions, the ResNet-18 classifier achieved AUCs of 0.964 and 0.952, respectively, on the test dataset. CONCLUSION: Deep learning classifiers effectively detect gonioscopic angle closure and PACD based on automated analysis of AS-OCT images. These methods could be used to automate clinical evaluations of the ACA and improve access to eye care in high-risk populations.
Authors: Monisha E Nongpiur; Mingguang He; Nishani Amerasinghe; David S Friedman; Wan-Ting Tay; Mani Baskaran; Scott D Smith; Tien Yin Wong; Tin Aung Journal: Ophthalmology Date: 2010-10-29 Impact factor: 12.079
Authors: Anas Z Abidin; Botao Deng; Adora M DSouza; Mahesh B Nagarajan; Paola Coan; Axel Wismüller Journal: Comput Biol Med Date: 2018-02-09 Impact factor: 4.589
Authors: Daniel Shu Wei Ting; Carol Yim-Lui Cheung; Gilbert Lim; Gavin Siew Wei Tan; Nguyen D Quang; Alfred Gan; Haslina Hamzah; Renata Garcia-Franco; Ian Yew San Yeo; Shu Yen Lee; Edmund Yick Mun Wong; Charumathi Sabanayagam; Mani Baskaran; Farah Ibrahim; Ngiap Chuan Tan; Eric A Finkelstein; Ecosse L Lamoureux; Ian Y Wong; Neil M Bressler; Sobha Sivaprasad; Rohit Varma; Jost B Jonas; Ming Guang He; Ching-Yu Cheng; Gemmy Chui Ming Cheung; Tin Aung; Wynne Hsu; Mong Li Lee; Tien Yin Wong Journal: JAMA Date: 2017-12-12 Impact factor: 56.272
Authors: Mingguang He; David S Friedman; Jian Ge; Wenyong Huang; Chenjin Jin; Pak Sang Lee; Peng T Khaw; Paul J Foster Journal: Ophthalmology Date: 2006-11-21 Impact factor: 12.079
Authors: Mohammed Rigi; Nicholas P Bell; David A Lee; Laura A Baker; Alice Z Chuang; Donna Nguyen; Vandana R Minnal; Robert M Feldman; Lauren S Blieden Journal: J Ophthalmol Date: 2016-11-20 Impact factor: 1.909
Authors: Ricardo J Cumba; Sunita Radhakrishnan; Nicholas P Bell; Kundandeep S Nagi; Alice Z Chuang; Shan C Lin; Kimberly A Mankiewicz; Robert M Feldman Journal: J Ophthalmol Date: 2012-11-01 Impact factor: 1.909
Authors: Alice Shen; Michael Chiang; Anmol A Pardeshi; Roberta McKean-Cowdin; Rohit Varma; Benjamin Y Xu Journal: Br J Ophthalmol Date: 2021-10-06 Impact factor: 4.638
Authors: Benjamin Y Xu; Jacob Lifton; Bruce Burkemper; Xuejuan Jiang; Anmol A Pardeshi; Sasan Moghimi; Grace M Richter; Roberta McKean-Cowdin; Rohit Varma Journal: Am J Ophthalmol Date: 2020-07-28 Impact factor: 5.258
Authors: Benjamin Y Xu; Siqi Liang; Anmol A Pardeshi; Jacob Lifton; Sasan Moghimi; Juan Pablo Lewinger; Rohit Varma Journal: Ophthalmol Glaucoma Date: 2020-09-14
Authors: An Ran Ran; Clement C Tham; Poemen P Chan; Ching-Yu Cheng; Yih-Chung Tham; Tyler Hyungtaek Rim; Carol Y Cheung Journal: Eye (Lond) Date: 2020-10-07 Impact factor: 3.775