Design and Development of Hybrid Optimization-Enabled Deep Learning Model for Myocardial Infarction

Abstract

Myocardial infarction is the most hazardous cardiovascular disease for humans; generally, it is acknowledged as a heart attack, which may result in death. Thus, rapid and precise detection of myocardial infarction is essential to reduce the mortality rate. This paper proposes the Taylor-enhanced invasive weed sine cosine optimization algorithm-based deep convolutional neural network (Taylor IIWSCOA-enabled DCNN) model to classify myocardial infarction. Here, the DCNN classifier is used to predict and categorize myocardial infarction, and the classifier is tuned by the Taylor IIWSCOA to attain superior efficiency. The Taylor IIWSCOA is designed by integrating SCA, IIWO approach, and the Taylor series. The proposed Taylor IIWSCOA-based DCNN approach outperforms other conventional approaches with an accuracy of 0.9412, sensitivity of 0.9535, and specificity of 0.9485.