Why the study?
Most machine learning algorithms for automatic ECG analysis are limited to initial training and datasets, failing to incorporate new expert-labeled signals that become available daily to enhance accuracy.
An active learning approach for ECG arrhythmia classification improves accuracy over traditional methods by utilizing newly labeled signals.
May enhance ECG arrhythmia classification; leaves open clinical adoption pending prospective validation.
Electrocardiography (ECG) is a widely used method for analyzing the condition of the heart. Due to the significant number of ECG signals that require analysis on a daily basis, the implementation of a machine learning algorithm as a medical assistant to automatically analyze these signals has become crucial. While several algorithms have been proposed for this purpose, most of them are limited to initial training and datasets. However, as a medical assistant system, new labeled signals provided by experts are available daily, which can be utilized to enhance accuracy and improve overall performance. Therefore, this paper introduces an active learning approach to address this issue. The proposed approach is evaluated using real data from multiple individuals and various scenarios to demonstrate its practicality and superior accuracy compared to traditional methods.
No takes yet. Share an insight, caveat, or question.
Darbani et al. (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: