Identifying the Abnormalities in Retinal Images Towards the Prediction of Cardiovascular Disease Using Deep Learning

One of the main causes of death worldwide is cardiovascular disease (CVD). It is crucial to discover CVDs early to avoid major consequences including heart attacks, strokes, and blindness. Certain alterations in the retinal microvasculature, including as constriction, tortuosity, and dilatation, have been scientifically shown in recent research to be precursors of CVDs. This is because the blood vessels in the retina are a direct extension of the brain, and the risk factors that affect the blood vessels in the brain and heart also affect the blood vessels in the retina. A dataset of retinal fundus images will be produced from a cohort of patients with and without CVDs with the calculation of Arteriolar to Venular Diameter Ratio (AVR) to assess the cardiovascular risk. This attempt will employ deep learning techniques such as VGG-16 and RESNET-50 to detect these retinal abnormalities and predict CVDs.