Clinical Decision Making in Cardiovascular Diseases Leveraging Deep Learning

Cardiovascular disease (CVD) remains one of the leading causes of morbidity and mortality worldwide, accounting for millions of deaths annually. Early diagnosis and accurate prediction of CVD are critical for effective prevention and management. This study introduces a hybrid deep learning model combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to improve CVD prognosis. The CNN component extracts spatial features from the data, while the LSTM component models temporal dependencies, enhancing the model's ability to detect patterns relevant to CVD risk. The proposed model achieved peak accuracies of 95.125% in training and 95% in validation with feature engineering. Without feature engineering, the model reached 92.25% in training accuracy and 94% in validation accuracy. These results demonstrate that the hybrid CNN-LSTM model outperforms traditional models, with feature engineering contributing to improved performance. Additionally, the integration of explainable AI helps identify the most important factors influencing CVD risk, providing valuable insights for the development of more effective screening tools. By combining feature engineering, deep learning, and explainable AI, this research advances the accuracy and interpretability of CVD prediction models' enhancing both precision and transparency in cardio-vascular care.