Why the study?
The authors sought to enable semi-supervised ECG models and facilitate the discovery of unknown arrhythmia subtypes and anomalous ECG signal events.
The release of the Icentia11K dataset provides a massive resource of 11,000 patients and 2 billion labeled beats to advance unsupervised representation learning for ECG arrhythmia detection and subtype discovery.
Current models show modest accuracy and should not yet change practice; leaves open the value of Icentia11K for advancing ECG arrhythmia detection.
We release the largest public ECG dataset of continuous raw signals for representation learning containing 11 thousand patients and 2 billion labelled beats. Our goal is to enable semi-supervised ECG models to be made as well as to discover unknown subtypes of arrhythmia and anomalous ECG signal events. To this end, we propose an unsupervised representation learning task, evaluated in a semi-supervised fashion. We provide a set of baselines for different feature extractors that can be built upon. Additionally, we perform qualitative evaluations on results from PCA embeddings, where we identify some clustering of known subtypes indicating the potential for representation learning in arrhythmia sub-type discovery.
No takes yet. Share an insight, caveat, or question.
Tan et al. (2019) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: