<![CDATA[Myocardial Infarction (MI) is a leading cause of global mortality, necessitating efficient diagnostic methods. This study develops a simplified one-dimensional Convolutional Neural Network (1D-CNN) model for classifying MI using 12-lead ECG signals from the PTB-XL dataset. The research focuses on reducing computational complexity by limiting convolutional layers while maintaining high accuracy. The proposed model processes ECG signals of varying lengths (600–1000 samples), identifying 700 samples as optimal, achieving an average accuracy of 96.18%, sensitivity of 82.84%, specificity of 97.63%, precision of 84.13%, and an F1-score of 82.68%. Leads V5 and V6 demonstrate superior performance in detecting MI, while other leads, such as I and AVL, require further optimization. By combining precise signal segmentation and an efficient CNN architecture, this model minimizes computational load without compromising performance, making it a strong candidate for real-time clinical applications. The findings highlight the importance of signal length optimization and simplified architecture in enhancing early MI detection.]]>
Copyrights © 2025
