A Hybrid Deep Learning Model for Heart Disease Risk Assessment

Deep Learning

Heart diseases remains a leading cause of mortality worldwide, necessitating effective risk assessment methodologies. In this study, we propose a comprehensive approach utilizing both clinical data and electrocardiogram (ECG) images to assess heart disease risk. Our dataset comprises clinical variables such as age, gender, blood pressure, cholesterol levels, along with ECG images depicting various cardiac arrhythmias including ventricular fibrillation (VFib), premature atrial contractions (PAC), premature ventricular contractions (PVC), left bundle branch block (LBBB), and right bundle branch block (RBBB). We employ various Machine Learning models and deep learning architectures, a hybrid model. For the analysis of ECG images, we leverage the power of convolutional neural networks (CNNs) with EfficientNetB0 architecture, known for its efficiency and effectiveness in image classification tasks. Through experimentation and evaluation, we demonstrate the performance of these models in predicting heart disease risk. This project provides insights into the potential of integrating clinical data and ECG images for enhanced risk assessment, contributing to the development of more accurate and personalized diagnostic tools in cardiology.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.