Application of an Anomaly Detection Model to Screen for Ocular Diseases Using Color Retinal Fundus Images: Design and Evaluation Study

Abstract

Background The supervised deep learning approach provides state-of-the-art performance in a variety of fundus image classification tasks, but it is not applicable for screening tasks with numerous or unknown disease types. The unsupervised anomaly detection (AD) approach, which needs only normal samples to develop a model, may be a workable and cost-saving method of screening for ocular diseases. Objective This study aimed to develop and evaluate an AD model for detecting ocular diseases on the basis of color fundus images. Methods A generative adversarial network–based AD method for detecting possible ocular diseases was developed and evaluated using 90,499 retinal fundus images derived from 4 large-scale real-world data sets. Four other independent external test sets were used for external testing and further analysis of the model’s performance in detecting 6 common ocular diseases (diabetic retinopathy [DR], glaucoma, cataract, age-related macular degeneration, hypertensive retinopathy [HR], and myopia), DR of different severity levels, and 36 categories of abnormal fundus images. The area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity of the model’s performance were calculated and presented. Results Our model achieved an AUC of 0.896 with 82.69% sensitivity and 82.63% specificity in detecting abnormal fundus images in the internal test set, and it achieved an AUC of 0.900 with 83.25% sensitivity and 85.19% specificity in 1 external proprietary data set. In the detection of 6 common ocular diseases, the AUCs for DR, glaucoma, cataract, AMD, HR, and myopia were 0.891, 0.916, 0.912, 0.867, 0.895, and 0.961, respectively. Moreover, the AD model had an AUC of 0.868 for detecting any DR, 0.908 for detecting referable DR, and 0.926 for detecting vision-threatening DR. Conclusions The AD approach achieved high sensitivity and specificity in detecting ocular diseases on the basis of fundus images, which implies that this model might be an efficient and economical tool for optimizing current clinical pathways for ophthalmologists. Future studies are required to evaluate the practical applicability of the AD approach in ocular disease screening.

Counteracting data bias and class imbalance – towards useful and reliable retinal disease recognition system

“…Despite lack of certain comparability, the authors attempted to compare AUC between presented model and the most recent studies. Although, AUC remains the most reliable measure of the learning algorithm performance 57 , only three studies reported this benchmark 11,14,16 . The model outperformed all three models in GL and AMD classi cation.…”

“…These diseases are prevalent in ageing populations, what make them suitable target for screening system 1,2,10 . Recently, there have been published several multiclass models that at least partially meet these conditions [11][12][13][14][15][16][17][18] . However, all these models had multiple limitations.…”

“…Merging multiple different datasets could even potentiate this issue. Most of the published studies 11,12,14,16,17 did not address the problem, what could in uence the results.…”

Detection of abnormal extraocular muscles in small datasets of computed tomography images using a three-dimensional variational autoencoder

“…In an ophthalmology study, an anomaly detection model was used to identify signs of ocular diseases in color retinal fundus images 16 . A generative adversarial network model was developed and evaluated using 90,499 retinal fundus images, and an area under the ROP curve value of 0.896, sensitivity of 82.69%, and specificity of 82.63% for detecting abnormal fundus images were achieved.…”

Progress on application of spatial epidemiology in ophthalmology

“…According to the first World Report on Vision issued by the World Health Organization in 2019, it is estimated that ~2.2 billion people suffer from vision impairment or blindness worldwide ( 1 ). Notably, this number will be still increased due to the growth of the global population ( 2 ).…”