SVM-Based Cardiac Risk Detection

Abstract

The leading causes of death around the world still include heart disease, affecting many countries, and personalized risk estimation, along with early detection, is needed to improve avoidance and treatment strategies. This project aims to create a clinically and diagnosis-relevant assessment of a person’s heart disease risk, utilizing machine learning techniques to predict personalized heart disease risk. The model implemented is a Support Vector Machine (SVM), and its basis is a large dataset that contains multiple medical parameters such as age, blood pressure, cholesterol level, and coronary artery calcification score.

A simple and secure web-based dashboard is designed for healthcare providers to enable registered members to log in and update or enter new data for active patient records, along with automated real-time risk forecasting. The backend module, which is built using Flask, handles all user data processing, user login authentication, and intercommunication with a systematic relational database holding user accounts, patient information, and the results from their corresponding evaluations and predictive analysis. The integration of machine learning and interactive applications leads to an innovative approach for early heart disease detection, which enhances clinical decisions and elevates patient care.