Evaluating Feature Attribution Methods for Electrocardiogram
This work addresses the need for reliable explanations in ECG-based cardiac arrhythmia detection, which is crucial for clinical practice, though it is incremental as it focuses on evaluating existing methods rather than proposing new ones.
The paper tackled the problem of evaluating feature attribution methods for explaining deep learning models in electrocardiogram (ECG) analysis, finding that Grad-CAM significantly outperforms other methods, with a large margin over the second-best method.
The performance of cardiac arrhythmia detection with electrocardiograms(ECGs) has been considerably improved since the introduction of deep learning models. In practice, the high performance alone is not sufficient and a proper explanation is also required. Recently, researchers have started adopting feature attribution methods to address this requirement, but it has been unclear which of the methods are appropriate for ECG. In this work, we identify and customize three evaluation metrics for feature attribution methods based on the characteristics of ECG: localization score, pointing game, and degradation score. Using the three evaluation metrics, we evaluate and analyze eleven widely-used feature attribution methods. We find that some of the feature attribution methods are much more adequate for explaining ECG, where Grad-CAM outperforms the second-best method by a large margin.