Application of artificial intelligence-based dual-modality analysis combining fundus photography and optical coherence tomography in diabetic retinopathy screening in a community hospital

Abstract

Background: To assess the feasibility and clinical utility of artificial intelligence (AI)-based screening for diabetic retinopathy (DR) and macular edema (ME) by combining fundus photos and optical coherence tomography (OCT) images in a community hospital.

Methods: Fundus photos and OCT images were taken for 600 diabetic patients in a community hospital. Ophthalmologists graded these fundus photos according to the International Clinical Diabetic Retinopathy (ICDR) Severity Scale as the ground truth. Two existing trained AI models were used to automatically classify the fundus images into DR grades according to ICDR, and to detect concomitant ME from OCT images, respectively. The criteria for referral were DR grades 2-4 and/or the presence of ME. The sensitivity and specificity of AI grading were evaluated. The number of referable DR cases confirmed by ophthalmologists and AI was calculated, respectively.

Results: DR was detected in 81 (13.5%) participants by ophthalmologists and in 94 (15.6%) by AI, and 45 (7.5%) and 53 (8.8%) participants were diagnosed with referable DR by ophthalmologists and by AI, respectively. The sensitivity, specificity and area under the curve (AUC) of AI for detecting DR were 91.67%, 96.92% and 0.944, respectively. For detecting referable DR, the sensitivity, specificity and AUC of AI were 97.78%, 98.38% and 0.981, respectively. ME was detected from OCT images in 49 (8.2%) participants by ophthalmologists and in 57 (9.5%) by AI, and the sensitivity, specificity and AUC of AI were 91.30%, 97.46% and 0.944, respectively. When combining fundus photos and OCT images, the number of referrals identified by ophthalmologists increased from 45 to 75 and from 53 to 85 by AI.

Conclusion: AI-based DR screening has high sensitivity and specificity and may feasibly improve the referral rate of community DR.

Keywords: Artificial intelligence; Deep learning; Diabetic retinopathy; Optical coherence tomography.

Fig. 1

OCT-Fundus-AI diagnosis results. a OCT B-scan with detection of retinal fluid (purple bounding box). b OCT B-scan with detection of the epiretinal membrane (yellow bounding box). c No obvious abnormalities on the fundus

Fig. 2

a Comparison of diabetic retinopathy (DR) grading between ophthalmologists and AI. b The confusion matrix for the DR detection. c Comparison of macular edema (ME) classifications between ophthalmologists and AI. d The confusion matrix for the ME detection

Fig. 3

Venn diagram showing the overlap comparison of the number of referrals between human and automated grading: a fundus photography, b OCT, and c fundus photography combined with OCT

Fig. 4

Flowchart of AI-based dual-modality DR screening algorithm

Fig. 5

Examples of classification heatmaps of different levels of DR

Fig. 6

Retinal abnormality detection results. Detected retinal exudates (white bounding box) and retinal fluid (purple bounding box)