Atherosclerosis disease prediction using Supervised Machine Learning Techniques
Oumaima Terrada, Bouchaib Cherradi, Abdelhadi Raihani, Omar Bouattane
Abstract
Atherosclerosis disease, also known as coronary artery disease (CAD) is the major reason to increase the mortality rate around the world. Indeed, there is a lack of improvement in the early diagnosis of cardiovascular diseases. Thus, doctors need a trustworthy system to minimize diagnostic errors and to avoid critically surgeries. This contribution is articulated around a Medical Decision Support System (MDSS) design for atherosclerosis, able to take precautionary steps using the patient's clinical parameters. This MDSS is based on supervised machine learning (ML) algorithms. We use Artificial Neural Network (ANN) and K-Nearest Neighbor (KNN) to predict patients with or without atherosclerosis disease in an established database. The system is validated on Cleveland heart disease, Hungarian, Switzerland, and Long Beach VA databases. The performance of the proposed system is evaluated using accuracy, sensitivity and specificity as well-known similarity measures. Our system outperforms currently similar published research.