Tiarnan D L Keenan1, Qingyu Chen2, Yifan Peng2, Amitha Domalpally3, Elvira Agrón1, Christopher K Hwang1, Alisa T Thavikulwat1, Debora H Lee1, Daniel Li2, Wai T Wong4, Zhiyong Lu5, Emily Y Chew6. 1. Division of Epidemiology and Clinical Applications, National Eye Institute, National Institutes of Health, Bethesda, Maryland. 2. National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda, Maryland. 3. Fundus Photograph Reading Center, University of Wisconsin-Madison, Madison, Wisconsin. 4. Division of Epidemiology and Clinical Applications, National Eye Institute, National Institutes of Health, Bethesda, Maryland; Section on Neuron-Glia Interactions in Retinal Disease, Laboratory of Retinal Cell and Molecular Biology, National Eye Institute, National Institutes of Health, Bethesda, Maryland. 5. National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health, Bethesda, Maryland. Electronic address: zhiyong.lu@nih.gov. 6. Division of Epidemiology and Clinical Applications, National Eye Institute, National Institutes of Health, Bethesda, Maryland. Electronic address: echew@nei.nih.gov.
Abstract
PURPOSE: To develop deep learning models for detecting reticular pseudodrusen (RPD) using fundus autofluorescence (FAF) images or, alternatively, color fundus photographs (CFP) in the context of age-related macular degeneration (AMD). DESIGN: Application of deep learning models to the Age-Related Eye Disease Study 2 (AREDS2) dataset. PARTICIPANTS: FAF and CFP images (n = 11 535) from 2450 AREDS2 participants. Gold standard labels from reading center grading of the FAF images were transferred to the corresponding CFP images. METHODS: A deep learning model was trained to detect RPD in eyes with intermediate to late AMD using FAF images (FAF model). Using label transfer from FAF to CFP images, a deep learning model was trained to detect RPD from CFP (CFP model). Performance was compared with 4 ophthalmologists using a random subset from the full test set. MAIN OUTCOME MEASURES: Area under the receiver operating characteristic curve (AUC), κ value, accuracy, and F1 score. RESULTS: The FAF model had an AUC of 0.939 (95% confidence interval [CI], 0.927-0.950), a κ value of 0.718 (95% CI, 0.685-0.751), and accuracy of 0.899 (95% CI, 0.887-0.911). The CFP model showed equivalent values of 0.832 (95% CI, 0.812-0.851), 0.470 (95% CI, 0.426-0.511), and 0.809 (95% CI, 0.793-0.825), respectively. The FAF model demonstrated superior performance to 4 ophthalmologists, showing a higher κ value of 0.789 (95% CI, 0.675-0.875) versus a range of 0.367 to 0.756 and higher accuracy of 0.937 (95% CI, 0.907-0.963) versus a range of 0.696 to 0.933. The CFP model demonstrated substantially superior performance to 4 ophthalmologists, showing a higher κ value of 0.471 (95% CI, 0.330-0.606) versus a range of 0.105 to 0.180 and higher accuracy of 0.844 (95% CI, 0.798-0.886) versus a range of 0.717 to 0.814. CONCLUSIONS: Deep learning-enabled automated detection of RPD presence from FAF images achieved a high level of accuracy, equal or superior to that of ophthalmologists. Automated RPD detection using CFP achieved a lower accuracy that still surpassed that of ophthalmologists. Deep learning models can assist, and even augment, the detection of this clinically important AMD-associated lesion. Published by Elsevier Inc.
PURPOSE: To develop deep learning models for detecting reticular pseudodrusen (RPD) using fundus autofluorescence (FAF) images or, alternatively, color fundus photographs (CFP) in the context of age-related macular degeneration (AMD). DESIGN: Application of deep learning models to the Age-Related Eye Disease Study 2 (AREDS2) dataset. PARTICIPANTS: FAF and CFP images (n = 11 535) from 2450 AREDS2 participants. Gold standard labels from reading center grading of the FAF images were transferred to the corresponding CFP images. METHODS: A deep learning model was trained to detect RPD in eyes with intermediate to late AMD using FAF images (FAF model). Using label transfer from FAF to CFP images, a deep learning model was trained to detect RPD from CFP (CFP model). Performance was compared with 4 ophthalmologists using a random subset from the full test set. MAIN OUTCOME MEASURES: Area under the receiver operating characteristic curve (AUC), κ value, accuracy, and F1 score. RESULTS: The FAF model had an AUC of 0.939 (95% confidence interval [CI], 0.927-0.950), a κ value of 0.718 (95% CI, 0.685-0.751), and accuracy of 0.899 (95% CI, 0.887-0.911). The CFP model showed equivalent values of 0.832 (95% CI, 0.812-0.851), 0.470 (95% CI, 0.426-0.511), and 0.809 (95% CI, 0.793-0.825), respectively. The FAF model demonstrated superior performance to 4 ophthalmologists, showing a higher κ value of 0.789 (95% CI, 0.675-0.875) versus a range of 0.367 to 0.756 and higher accuracy of 0.937 (95% CI, 0.907-0.963) versus a range of 0.696 to 0.933. The CFP model demonstrated substantially superior performance to 4 ophthalmologists, showing a higher κ value of 0.471 (95% CI, 0.330-0.606) versus a range of 0.105 to 0.180 and higher accuracy of 0.844 (95% CI, 0.798-0.886) versus a range of 0.717 to 0.814. CONCLUSIONS: Deep learning-enabled automated detection of RPD presence from FAF images achieved a high level of accuracy, equal or superior to that of ophthalmologists. Automated RPD detection using CFP achieved a lower accuracy that still surpassed that of ophthalmologists. Deep learning models can assist, and even augment, the detection of this clinically important AMD-associated lesion. Published by Elsevier Inc.
Authors: Tiarnan D L Keenan; Qingyu Chen; Elvira Agrón; Yih-Chung Tham; Jocelyn Hui Lin Goh; Xiaofeng Lei; Yi Pin Ng; Yong Liu; Xinxing Xu; Ching-Yu Cheng; Mukharram M Bikbov; Jost B Jonas; Sanjeeb Bhandari; Geoffrey K Broadhead; Marcus H Colyer; Jonathan Corsini; Chantal Cousineau-Krieger; William Gensheimer; David Grasic; Tania Lamba; M Teresa Magone; Michele Maiberger; Arnold Oshinsky; Boonkit Purt; Soo Y Shin; Alisa T Thavikulwat; Zhiyong Lu; Emily Y Chew Journal: Ophthalmology Date: 2022-01-03 Impact factor: 14.277
Authors: Mingquan Lin; Bojian Hou; Lei Liu; Mae Gordon; Michael Kass; Fei Wang; Sarah H Van Tassel; Yifan Peng Journal: Sci Rep Date: 2022-08-18 Impact factor: 4.996
Authors: Tiarnan D L Keenan; Neal L Oden; Elvira Agrón; Traci E Clemons; Alice Henning; Lars G Fritsche; Wai T Wong; Emily Y Chew Journal: Ophthalmol Retina Date: 2021-07-26