Deep Learning Approach for Automated Detection of Myopic Maculopathy and Pathologic Myopia in Fundus Images.

Abstract Purpose To determine whether eyes with pathologic myopia can be identified and whether each type of myopic maculopathy lesion in fundus photographs can be diagnosed by deep learning (DL) algorithms. Design A DL algorithm was developed to recognize myopic maculopathy features and to automatically categorize the myopic maculopathy. Subjects We examined 7020 fundus images from 4432 highly myopic eyes obtained from the Advanced Clinical Center for Myopia. Methods DL algorithms were developed to recognize the key features of myopic maculopathy with 5176 fundus images of 2588 highly myopic eyes. These algorithms were also used to develop a META-PM categorizing system (META-PM CS) by adding a specific processing layer. Models and system were evaluated by 1844 fundus images of 1844 highly myopic eyes. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve, the sensitivity, and specificity were used to determine the performance of each DL algorithm. The rate of correct predictions was used to determine the performance of the META-PM CS. Main Outcome Measures Four trained DL models were able to recognize the lesions of myopic maculopathy accurately with high sensitivity and specificity. The META-PM CS also had a high accuracy and was qualified to be used in a semi-automated way during screening for myopic maculopathy in highly myopic eyes. Results The sensitivity of the DL models was 84.44% for diffuse atrophy, 87.22% for patchy atrophy, 85.10% for macular atrophy, and 37.07% for choroidal neovascularization, and the AUC values with 95% CI of 0.970 (0.966-0.974), 0.978 (0.967-0.987), 0.982 (0.971-0.994) and 0.881 (0.854-0.902), respectively. The rate of total correct predictions from META-PM CS was 87.53%, with rates of 90.18%, 95.28%, 97.50%, and 91.14% respectively for each type of lesion. The META-PM CS had an overall rate of 92.08% in detecting pathologic myopia correctly which was defined as having myopic maculopathy equal to or more serious than diffuse atrophy. Conclusions The novel DL models and systems can achieve high sensitivity and specificity in identifying the different types of lesions of myopic maculopathy. These results will assist in the screening for pathologic myopia and subsequent protection of patients against low vision and blindness caused by myopic maculopathy.