Deep learning algorithms have been used to detect diabetic retinopathy (DR) with specialist-level accuracy. This study aims to validate one such algorithm on a large-scale clinical population, and compare the algorithm performance with that of human graders. A total of 25,326 gradable retinal images of patients with diabetes from the community-based, nationwide screening program of DR in Thailand were analyzed for DR severity and referable diabetic macular edema (DME). Grades adjudicated by a panel of international retinal specialists served as the reference standard. Relative to human graders, for detecting referable DR (moderate NPDR or worse), the deep learning algorithm had significantly higher sensitivity (0.97 vs. 0.74, p < 0.001), and a slightly lower specificity (0.96 vs. 0.98, p < 0.001). Higher sensitivity of the algorithm was also observed for each of the categories of severe or worse NPDR, PDR, and DME ( p < 0.001 for all comparisons). The quadratic-weighted kappa for determination of DR severity levels by the algorithm and human graders was 0.85 and 0.78 respectively ( p < 0.001 for the difference). Across different severity levels of DR for determining referable disease, deep learning significantly reduced the false negative rate (by 23%) at the cost of slightly higher false positive rates (2%). Deep learning algorithms may serve as a valuable tool for DR screening.
IMPORTANCE Diagnosing diabetic macular edema (DME) from monocular fundus photography vs optical coherence tomography (OCT) central subfield thickness (CST) can yield different prevalence rates for DME. Epidemiologic studies and telemedicine screening typically use monocular fundus photography, while treatment of DME uses OCT CST. OBJECTIVE To compare DME prevalence from monocular fundus photography and OCT. DESIGN, SETTING, AND PARTICIPANTS Retrospective cross-sectional study of DME grading based on monocular fundus photographs and OCT images obtained from patients with diabetic retinopathy at a single visit between July 1, 2011, and June 30, 2014, at a university-based practice and analyzed between July 30, 2014, and May 29, 2015. Presence of DME, including clinically significant macular edema (CSME), on monocular fundus photographs used definitions from the Multi-Ethnic Study of Atherosclerosis (MESA) and the National Health and Nutrition Examination Survey (NHANES). Presence of DME on OCT used Diabetic Retinopathy Clinical Research Network eligibility criteria thresholds of CST for trials evaluating anti-vascular endothelial growth factor treatments. MAIN OUTCOMES AND MEASURES Prevalence of DME based on monocular fundus photographs or OCT. RESULTS A total of 246 eyes of 158 participants (mean [SD] age, 65.0 [11.9] years; 48.7% women; 60.8% white) were included. Among the 246 eyes, the prevalences of DME (61.4%) and CSME (48.5%) based on MESA definitions for monocular fundus photographs were greater than the DME prevalence based on OCT (21.1%) by 40.2% (95% CI, 32.8%-47.7%; P < .001) and 27.2% (95% CI, 19.2%-35.3%; P < .001), respectively. Using NHANES definitions, DME and CSME prevalences from monocular fundus photographs (28.5% and 21.0%, respectively) approximated the DME prevalence from OCT (21.1%). However, among eyes without DME on OCT, 58.2% (95% CI, 51.0%-65.3%) and 18.0% (95% CI, 12.9%-24.2%) were diagnosed as having DME on monocular fundus photographs using MESA and NHANES definitions, respectively, including 47.0% (95% CI, 39.7%-54.5%) and 10.3% (95% CI, 6.3%-15.7%), respectively, with CSME. Among eyes with DME on OCT, 26.9% (95% CI, 15.6%-41.0%) and 32.7% (95% CI, 20.3%-47.1%) were not diagnosed as having either DME or CSME on monocular fundus photographs using MESA and NHANES definitions, respectively. CONCLUSIONS AND RELEVANCE These data suggest that many eyes diagnosed as having DME or CSME on monocular fundus photographs have no DME based on OCT CST, while many eyes diagnosed as not having DME or CSME on monocular fundus photographs have DME on OCT. While limited to 1 clinical practice, caution is suggested when extrapolating prevalence of eyes that may benefit from anti-vascular endothelial growth factor therapy based on epidemiologic surveys using photographs to diagnose DME.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.