Search papers, labs, and topics across Lattice.
This paper explores transformer-based ECG classification using features extracted via Koopman operator theory (EDMD) and wavelet transforms. They compare these features for binary (Normal vs. Non-normal) and four-class ECG classification tasks, finding that Koopman features outperform wavelet features in the four-class setting when used with transformers. The key result is that a carefully tuned radial basis function dictionary for EDMD significantly improves performance compared to wavelet-only and hybrid approaches, demonstrating the potential of Koopman-based feature learning.
Forget wavelets, transformers with Koopman operator-derived features unlock superior ECG classification, especially in complex multi-class scenarios.
Electrocardiogram (ECG) analysis is vital for detecting cardiac abnormalities, yet robust automated classification is challenging due to the complexity and variability of physiological signals. In this work, we investigate transformer-based ECG classification using features derived from the Koopman operator and wavelet transforms. Two tasks are studied: (1) binary classification (Normal vs. Non-normal), and (2) four-class classification (Normal, Atrial Fibrillation, Ventricular Arrhythmia, Block). We use Extended Dynamic Mode Decomposition (EDMD) to approximate the Koopman operator. Our results show that wavelet features excel in binary classification, while Koopman features, when paired with transformers, achieve superior performance in the four-class setting. A simple hybrid of Koopman and wavelet features does not improve accuracy. However, selecting an appropriate EDMD dictionary -- specifically a radial basis function dictionary with tuned parameters -- yields significant gains, surpassing the wavelet-only baseline and the hybrid wavelet-Koopman system. We also present a Koopman-based reconstruction analysis for interpretable insights into the learned dynamics and compare against a recurrent neural network baseline. Overall, our findings demonstrate the effectiveness of Koopman-based feature learning with transformers and highlight promising directions for integrating dynamical systems theory into time-series classification.
