A Hybrid Deep Learning Framework for Heart Disease Risk Assessment

cardiovascular diseases (CVDs), including coronary artery disease (CAD), remain the leading cause of mortality worldwide, necessitating early and accurate diagnostic methods. Recent advances in deep learning have enabled the integration of multimodal data to enhance predictive performance in medical diagnostics. The primary objective of this research is to develop a deep learning-based framework that leverages retinal images and corresponding pathological data to predict CAD, addressing the gap in multimodal learning for cardiovascular risk assessment. Traditional CAD diagnostic approaches rely on invasive and expensive procedures, whereas this study aims to provide a non-invasive, cost-effective alternative by utilizing fundus imaging and clinical parameters. The proposed methodology employs a Convolutional Neural Network (CNN) to extract spatial features from retinal images and a Recurrent Neural Network (RNN) to model temporal dependencies in associated pathological data. These modalities are fused through a hybrid deep learning architecture, ensuring an effective combination of visual and clinical information. The novelty of the proposed framework lies in its ability to integrate CNN-extracted retinal biomarkers with sequential RNN-based pathological patterns, enhancing CAD prediction accuracy. Advanced optimization techniques and feature fusion strategies further contribute to the robustness of the model. Experimental results demonstrate that the proposed approach significantly improves predictive performance compared to conventional machine learning methods. The integrated CNN-RNN model achieves high sensitivity and specificity, outperforming existing unimodal and traditional approaches.