Classification of ECG Signals with Different Lead Systems Using AutoML

The PhysioNet 2021 challenge asks participants to develop automated techniques for classifying cardiac abnormalities (CA) from both 12-lead electrocardiogram (ECG) and reduced-lead settings. We investigated on the feasibility of applying Automated Machine Learning (AutoML) approaches to build ECG classifiers. Standard ECG preprocessing was applied to the ECG (filtering and resampling), Three different AutoML frameworks were executed on the 88,000+ ECGs made available by the challenge organizers. The optimal ML pipeline was found by the Au-toML frameworks. We finally assessed the frameworks' classification performance, the effect of the number of employed leads, and the effect of extending the frameworks training time. The classifiers of our team “BiSPLab” received scores of 0.30, 0.29, 0.28, 0.26, 0.23 (ranked 27th, 29th, 28th, 29th, 28th out of 39 teams) for the 12-lead, 6-lead, 4-lead, 3-lead, and 2-lead versions of the hidden test set with the Challenge evaluation metric. The Au-toML frameworks showed comparable performance. Significantly extending the training time seemed to not improve the test score. AutoML showed promising performance on the test set, suggesting their potential for classification of CA. Future works are towards testing further Au-toML approaches, and better determining the impact of the available training time on the classification performance.