ECG-Lens Study Compares Machine Learning and Deep Learning Models for ECG Signal Classification

A research study published on arXiv (ID: 2604.15822v1) benchmarks multiple machine learning and deep learning models for automated electrocardiogram (ECG) signal classification using the PTB-XL dataset. The PTB-XL dataset contains 12-lead ECG recordings from both normal patients and individuals with various cardiac conditions. Three traditional machine learning algorithms—Decision Tree Classifier, Random Forest Classifier, and Logistic Regression—were compared against three deep learning architectures: a Simple Convolutional Neural Network (CNN), a Long Short-Term Memory (LSTM) network, and a Complex CNN model named ECGLens. Deep learning models were trained directly on raw ECG signals, enabling automatic extraction of discriminative features without manual feature engineering. To enhance model performance and increase training sample diversity, data augmentation was implemented using the Stationary Wavelet Transform (SWT), which preserves essential characteristics of ECG signals. Model evaluation employed multiple metrics to assess classification accuracy and robustness. This automated classification approach serves as a diagnostic and monitoring tool for cardiovascular diseases. The study represents a cross-disciplinary application of computational methods to medical data analysis.