This study evaluates machine learning and deep learning models for classifying ECG signals from the PTB-XL dataset. It compares six models, including Random Forest, Logistic Regression, CNN, LSTM, and a complex CNN called ECG-Lense. The ECG-Lense model achieved the highest accuracy (80%) and ROC-AUC (90%) using raw signals and data augmentation. Findings support automated ECG analysis for improved diagnosis and future cardiac research.
