Abstract
•This review gives a comprehensive view on stat-of-the-art deep learning based methods related to diabetic retinopathy diagnosis and will help researchers to conduct further research on this problem.•The survey gives an overview of different diabetic retinopathy biomarkers and lesions and different tasks related to diagnosis. Also we have given an overview of datasets, which have been developed for research on diagnosis and commonly used performance metrics.•Various bibliographic reference sources are used to explore the diverse applications of deep learning algorithms in the diabetic retinopathy diagnosis where most of which appeared recently.•It covers deep learning based methods, which have been proposed for retinal blood vessels segmentation, OD detection and segmentation, detection of various lesions such as EXs, MAs, HMs and referable diabetic retinopathy.•The review mentioned deep learning based methods in diabetic retinopathy and discussed their overall performance, gaps and future directions. Also it gives comparison with stat-of-the-art traditional methods based on hand-engineered features.
Diabetic retinopathy (DR) results in vision loss if not treated early. A computer-aided diagnosis (CAD) system based on retinal fundus images is an efficient and effective method for early DR diagnosis and assisting experts. A computer-aided diagnosis (CAD) system involves various stages like detection, segmentation and classification of lesions in fundus images. Many traditional machine-learning (ML) techniques based on hand-engineered features have been introduced. The recent emergence of deep learning (DL) and its decisive victory over traditional ML methods for various applications motivated the researchers to employ it for DR diagnosis, and many deep-learning-based methods have been introduced. In this paper, we review these methods, highlighting their pros and cons. In addition, we point out the challenges to be addressed in designing and learning about efficient, effective and robust deep-learning algorithms for various problems in DR diagnosis and draw attention to directions for future research.