Enhanced Machine Learning-Assisted Convolutional Neural Network for Heart Disease Prediction

Abstract

The accurate and timely diagnosis of heart disease remains a critical challenge in modern healthcare. Recent advancements in deep learning have paved the way for intelligent diagnostic systems that enhance medical decision-making. This study introduces an Enhanced Deep Learning-Assisted Convolutional Neural Network (EDCNN) for heart disease prediction, leveraging the Internet of Medical Things (IoMT) to enable real-time and remote diagnostics. The proposed model integrates a multi-layer perceptron (MLP) framework with optimized regularization techniques, ensuring robust feature extraction and classification. The system’s performance is evaluated using both comprehensive and reduced feature sets to analyze the trade-off between computational efficiency and diagnostic accuracy. Experimental results demonstrate that EDCNN outperforms traditional models such as Artificial Neural Networks (ANN), Deep Neural Networks (DNN), and Recurrent Neural Networks (RNN) in terms of precision, recall, and overall predictive accuracy. Implemented on a cloud-based IoMT platform, the model facilitates seamless access to diagnostic insights, supporting healthcare professionals worldwide. Comparative analysis indicates that fine-tuning EDCNN’s hyperparameters enables it to achieve a remarkable precision rate of 99.1%, reinforcing its potential as a reliable and efficient tool for heart disease prognosis.