Early Detection of Genetically Influenced Cardiovascular Disease Using Hybrid CNN-LSTM on ECG Data

Abstract

Cardiovascular disease (CVD) is a leading cause of morbidity and mortality worldwide, and early detection is critical for improving patient outcomes. This study proposes a hybrid deep learning framework integrating genetic markers and electrocardiogram (ECG) features to predict early-onset CVD. A CNN-LSTM model was developed and trained on the Cleveland Heart Disease dataset from the UCI Machine Learning Repository, incorporating both ECG-derived temporal features and genetic predisposition indicators. The model achieved an accuracy of 92.5%, precision of 91.2%, recall of 90.8%, F1-score of 91.0%, and an AUC-ROC of 0.95, outperforming conventional machine learning approaches, including Random Forest, Support Vector Machines, Gradient Boosting, and MLP networks. Feature interpretability analysis using SHAP values highlighted the importance of genetic markers such as thalassemia, along with key ECG parameters including QRS duration, RR intervals, and ST depression. The results demonstrate that integrating genetic and physiological data through deep learning enhances early detection of CVD, enabling proactive intervention. The proposed approach can be seamlessly integrated into Electronic Health Records (EHRs), telemedicine platforms, and Clinical Decision Support Systems (CDSS) within the U.S. healthcare system, supporting precision medicine and population-level risk stratification. This study underscores the potential of AI-driven predictive models in transforming cardiovascular healthcare by providing personalized, timely, and accurate risk assessments.