Robust and Interactive Detection System for Cardiovascular Disease using Artificial Intelligence

Abstract

Cardiovascular disease is a generic term that encompasses a variety of illnesses affecting the heart and blood vessels. Cardiovascular disease prediction is challenging due to the presence of several parameters. Existing cardiac parameter measurements apparatus serve as standalone measurement devices and cannot convey useful predictions. Coupling measurement devices with an intelligence system by incorporating user interactivity and a robust classification algorithm will yield a better holistic system. The proposed system attempts to detect cardiovascular disease at early stages with decent accuracy. The critical parameters necessary for this disease prediction are extracted by pulse oximeter sensor in real-time and additional non-sensor parameters are acquired from the patient to precisely detect the occurrence of cardiovascular disease. The degree of interactivity in the system possesses a smooth user experience throughout the detection process with the help of a web interface and email notifications. Through appropriate intimation, a patient can be saved from a situation of increased complexity to avoid disastrous consequences. The results verify that the optimized Decision Tree classifier has achieved the highest testing accuracy of 99.79% in minimal run time complexity. The designed system will timely aid cardio respiratory patients for their welfare and suggests a unique way to avoid the physical presence of the patient in the hospital for consultation.