Key result
CNN-DVIT model beats latest transformer algorithms for automatic ECG arrhythmia detection with ~83% F1 score.
Why the study?
Current transformer-based neural networks have limited performance in detecting arrhythmias from multi-lead ECG recordings.
Population
12-lead ECG recordings with varied lengths from CPSC-2018
Comparison
CNN-DVIT model vs latest transformer-based ECG classification algorithms
Design
Model development and validation study
Authors
Loading...
Supports hybrid CNN-transformer ECG models in research; leaves open clinical adoption pending prospective validation.
The proposed CNN-DVIT model demonstrates high accuracy in automatic arrhythmia classification from 12-lead ECGs, potentially aiding clinical diagnosis.
Dong et al. (2023) studied Arrhythmia. CNN-DVIT model vs. Latest transformer-based ECG classification algorithms was evaluated on F1 score. The CNN-DVIT model achieved an F1 score of 82.9% in the CPSC-2018 dataset, outperforming the latest transformer-based ECG classification algorithms for automatic arrhythmia detection.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: