Development and Validation of Deep Learning Models for Screening Multiple Abnormal Findings in Retinal Fundus Images

Abstract

Purpose: To develop and evaluate deep learning models that screen multiple abnormal findings in retinal fundus images. Design: Cross-sectional study. Participants: For the development and testing of deep learning models, 309 786 readings from 103 262 images were used. Two additional external datasets (the Indian Diabetic Retinopathy Image Dataset and e-ophtha) were used for testing. A third external dataset (Messidor) was used for comparison of the models with human experts. Methods: Macula-centered retinal fundus images from the Seoul National University Bundang Hospital Retina Image Archive, obtained at the health screening center and ophthalmology outpatient clinic at Seoul National University Bundang Hospital, were assessed for 12 major findings (hemorrhage, hard exudate, cottonwool patch, drusen, membrane, macular hole, myelinated nerve fiber, chorioretinal atrophy or scar, any vascular abnormality, retinal nerve fiber layer defect, glaucomatous disc change, and nonglaucomatous disc change) with their regional information using deep learning algorithms. Main Outcome Measures: Area under the receiver operating characteristic curve and sensitivity and specificity of the deep learning algorithms at the highest harmonic mean were evaluated and compared with the performance of retina specialists, and visualization of the lesions was qualitatively analyzed. Results: Areas under the receiver operating characteristic curves for all findings were high at 96.2% to 99.9% when tested in the in-house dataset. Lesion heatmaps highlight salient regions effectively in various findings. Areas under the receiver operating characteristic curves for diabetic retinopathy-related findings tested in the Indian Diabetic Retinopathy Image Dataset and e-ophtha dataset were 94.7% to 98.0%. The model demonstrated a performance that rivaled that of human experts, especially in the detection of hemorrhage, hard exudate, membrane, macular hole, myelinated nerve fiber, and glaucomatous disc change. Conclusions: Our deep learning algorithms with region guidance showed reliable performance for detection of multiple findings in macula-centered retinal fundus images. These interpretable, as well as reliable, classification outputs open the possibility for clinical use as an automated screening system for retinal fundus images.

Effects of Hypertension, Diabetes, and Smoking on Age and Sex Prediction from Retinal Fundus Images

“…Dataset Organisation. We used the retinal fundus images from the Seoul National University Bundang Hospital Retina Image Archive (SBRIA) after de-identification except the age, sex, and underlying diseases at the study date; details are described in our previous study 26,30 . We included 412,026 retinal fundus images from 155,449 participants obtained at the health promotion centre in Seoul National University Bundang Hospital (SNUBH) between June 1st, 2003, and June 30th, 2016, in which detailed information regarding the presence of hypertension, DM, and the smoking status was presented.…”

“…In particular, as retinal fundus images provide high-resolution in-vivo images of retinal vessels and parenchyma without any invasive procedure 15 , retinal fundus images have been used to detect target organ damage in vascular diseases (e.g., hypertension and diabetes mellitus [DM]), to screen retinal and optic disc diseases (e.g., age-related macular degeneration and glaucoma), and to predict cerebral/cardiovascular diseases [16][17][18][19][20][21] . Recently, deep neural networks have revolutionised the field of medical image analysis including retinal fundus images; deep-learning algorithms presented discriminative performances comparable to those of an ophthalmologist in diagnosing diabetic retinopathy, age-related macular degeneration, and glaucoma, as well as in predicting age, sex, and presence of cardiovascular diseases [22][23][24][25][26] . Previous study reported highly accurate results of a mean absolute error of 3.26 years for age prediction and receiver operating characteristic curve (AUC) of 0.97 for gender prediction 25 , however, there were several limitations including a low proportion of Asian, a lack of analysis of the differences in accuracy with age, and a lack of discussion of which parts of the fundus image were used for age prediction.…”

Development and Evaluation of a Deep Learning System for Screening Retinal Hemorrhage Based on Ultra-Widefield Fundus Images

“…Second, although DL systems developed based on traditional fundus images showed reliable performance for the detection of retinopathies, missed diagnoses were inevitable due to the limited visible scope of these fundus images. 16,26 Our study developed a robust DL model for detecting RH based on UWF images, the visible scope of which was five times larger than that of traditional fundus images. Thus, our system minimizes missed diagnoses of RH.…”

Machine learning applied to retinal image processing for glaucoma detection: review and perspective

“…Fundus imaging allows the identification of the main ocular structures, such as the optic disc (OD), optic disc cup (OD-cup), macula region [12], fovea, [13] and blood vessels [14]. This test may also detect abnormal conditions, including microaneurysms, bleeding, exudates, and cotton wool spots [15].…”