Automated Computationally Intelligent Methods for Ocular Vessel Segmentation and Disease Detection: A Review

Ocular diseases are eventually increasing these days that cause partial or complete vision loss even at an early age, the prominent reason behind this is cardiovascular diseases that cause diabetic retinopathy (DR), hypertensive retinopathy(HR), Glaucoma, and abnormal lesions. For early diagnosis of these diseases, we need to analyze the internal anatomical structure of the retina especially the morphology of vasculature. Identification of subtle abnormalities in blood vessels is required for accurate diagnosis of different ocular diseases as blood vessels are the key features of the retina. Many researchers have elucidated different traditional as well as automated machine learning and deep learning-based methodologies for vessel segmentation and disease detection. In the majority of recent techniques, the performance parameters are improved time and again. The advancement in automated techniques can propel early diagnosis so an extensive study is needed to find the gaps. A systematic review of various existing methodologies for vessel segmentation and ocular disease detection and classification is presented in this paper that gives a comprehensive idea about the recent trends related to this subject. A comparative analysis of various methodologies that have been used so far for computer-aided diagnosis of different ocular disease is presented.