Why the study?
Continuously acquired biosignals from patient monitors contain significant amounts of unusable data, creating a need for algorithms to automatically classify ECG data quality.
Does a two-step machine learning approach improve the accuracy of ECG data quality classification compared to a one-step approach in imbalanced datasets?
Population
31,127 training/validation and 9779 external test twenty-s ECG segments of 250 Hz
Comparison
Two-step binary sequential approaches vs one-step three-class approaches using RF and 2D CNN
Design
Algorithm development and external validation study
Authors
Loading...
May improve ECG quality classification in imbalanced datasets; hypothesis-generating for clinical workflows.
Does a two-step machine learning approach improve the accuracy of ECG data quality classification compared to a one-step approach in imbalanced datasets?
A two-step machine learning approach provides more robust performance than a one-step approach for classifying ECG data quality in highly imbalanced datasets.
Kim et al. (2023) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: