Automated Cardiovascular Disease Detection with Hybrid Machine Approach

Abstract

Cardiovascular ailment (CVD) is a main reason of death and disability globally, posing tremendous challenges for early detection and prevention. A heart condition or the accumulation of fatty deposits inside the arteries, is frequently linked to CVD and raises the risk of blood clots and other consequences. Predicting and diagnosing CVD is critical to mitigating its effect and enhancing patient results. With the appearance of massive data in healthcare, large volumes of patient records are available that may assist in identifying early warning symptoms of coronary heart disorder. This project aims to generate a hybrid machine learning technique which leverages deep learning and traditional machine learning algorithms to automatically locate cardiovascular illnesses. By integrating more than one dataset, we intend to upgrade the model's overall performance in diagnosing coronary heart situations. Numerous machine learning classification models, consisting of Random Forest and Gradient Boosting, might be employed and in comparison, to evaluate their effectiveness in detecting CVD. Each model can be evaluated using overall performance indicators like precision, accuracy, recall, and F1-Score. In previous research, the Random Forest model did 94% accuracy in coronary heart disorder detection, at the same time as the Gradient Boosting model established a balanced overall performance through accuracy, precision, recall, and F1-score with 73% each in detecting cardiovascular diseases via this hybrid technique. We aim to enhance prediction accuracy and offer a greater reliable tool for early CVD detection, in the long run enhancing affected person care and saving lives.