Diabetic Retinopathy Early Detection Based on OCT and OCTA Feature Fusion

Abstract

Diabetic retinopathy (DR) is a leading cause of vision loss in adults between 20 and 74 years. It is one of the significant causes of blindness worldwide. It affects the blood vessels in the retina as a complication of diabetes. Early detection of DR could reduce the severity of the disease and help ophthalmologists in treating and investigating it more efficiently. In this paper, a computer-aided diagnosis (CAD) system was developed for the detection of the early signs of DR using both optical coherence tomography angiography (OCTA) and optical coherence tomography (OCT) scans. This system was able to segment the blood vessels from OCTA images and twelve retinal layers from OCT images. Then, five different features were extracted from both segmented images. Finally, these features are fed to a support vector machines (SVM) to generate a CAD system for the early detection of DR. Using 2-fold cross-validation, this system achieved an average accuracy of 97%.