Early Detection and Prognostic Assessment of Ischemic Heart Diseases Based on Multimodal Artificial Intelligence Algorithms

Abstract

Cardiovascular diseases remain one of the leading causes of mortality worldwide, among which myocardial infarction is the most critical condition requiring rapid and accurate diagnosis. In this study, a deep learning approach based on convolutional neural networks (CNNs) was developed for the detection of myocardial infarction through automated analysis of electrocardiogram (ECG) signals. The PTB Diagnostic ECG Database was selected as the primary dataset. ECG signal preprocessing included baseline wander correction, noise filtering, normalization, and segmentation using the Pan-Tompkins algorithm. Based on the labeled ECG segments, a one-dimensional CNN (1D-CNN) architecture was designed and trained. On the test dataset, the proposed model achieved an accuracy of 93.1%, sensitivity of 91.4%, specificity of 95.2%, and an F1-score of 92.3%. The results demonstrate that CNN-based models can effectively identify ECG patterns characteristic of myocardial infarction and serve as a reliable and rapid tool for early diagnosis. This study confirms the significant potential of deep learning algorithms in improving cardiac disease diagnostic systems and highlights their feasibility for real-time integration into clinical or remote monitoring environments.