HEART ATTACK PREDICTION USING FUNDUS IMAGE BY APPLYING DEEP LEARNING

Abstract

Heart diseases are the most cause of passing around the world. Therefore, early detection and risk assessment of cardiovascular diseases are important for effective prevention and timely intervention. Current strategy for analyze is depending on investigation of persistent wellbeing records (blood test, ECG, stress test), history of understanding and doctor's instinct, which is time-consuming and detailed. To overcome this challenge, we propose a progressed approach for heart assault location based on retinal pictures utilizing diverse AI techniques. In this project, we developed an “AI program” can scan the image of patient’s eye that is retina and by studying the power of the blood vessels present in eye that cater for the retina find signs that indicate possibility of a heart attack. It also identifies range of life span, highest value of blood pressure that is systolic, lowest value of blood pressure that is diastolic, body mass index as well as hemoglobin level of person. The purpose to develop this design is to identify new disease detection or monitoring approaches which are less obtrusive, more precise, cheaper and more promptly accessible.