Key result
A domain-adversarial learning strategy based on the shifted window transformer (DA-ST) achieved an average atrial fibrillation detection accuracy of 98.67% on wearable ECG recordings, outperforming models without domain adaptation.
Why the study?
AF detectors can suffer degraded performance on wearable ECGs due to significant individual differences and interference from other arrhythmias.
Does a domain-adversarial learning strategy improve the accuracy of atrial fibrillation detection in dynamic ECGs with significant individual differences?
Does a domain-adversarial learning strategy improve the accuracy of atrial fibrillation detection in dynamic ECGs with significant individual differences?
A domain-adversarial learning strategy significantly improves the accuracy of deep learning models for detecting atrial fibrillation in dynamic wearable ECGs by mitigating the impact of individual differences and premature beats.
Answers are grounded in this paper and cite their sources.
May improve wearable AF detection amid inter-individual variability; extends domain-adversarial methods but remains hypothesis-generating.
Atrial fibrillation (AF) is an insidious disease. Many long-term wearable electrocardiogram (ECG) monitoring devices have been used to monitor AF. The accuracy of detectors used to classify AF/sinus rhythm is already very high on the public database. Due to the significant individual differences and interference from other arrhythmias (e.g., premature beats), the performance of the developed AF detectors can degrade when tested on wearable ECGs. To tackle this, we proposed to use a domain-adversarial (DA) learning strategy to minimize feature distribution between the annotated public ECG database (the MIT-BIH AF database) and unlabeled dynamic ECG recordings to improve AF recognition accuracy. DA algorithms based on the shifted window transformer (DA-ST) and residual neural network (DA-RN) were proposed and validated on the 2021 China Physiological Signal Challenge (CPSC) database including four datasets. The accuracies were 93.85%, 89.78%, 91.93%, and 87.35% on the four datasets when using DA-ST. The corresponding results were 96.67%, 92.25%, 90.58%, and 89.46% when utilizing DA-RN. Importantly, these results demonstrated superior performance compared to the results obtained without DA. The proposed method was validated on 12 wearable long-term recordings, consisting of four recordings with premature beats, four recordings of AF with premature beats, two recordings of sinus rhythms, and two recordings of AF. The average results were 98.67% (DA-ST) and 97.89% (DA-RN), proving that the proposed method could provide reliable AF detection for dynamic ECG recordings with significant individual differences.
No takes yet. Share an insight, caveat, or question.
Ma et al. (2023) studied Atrial fibrillation. Domain-adversarial learning strategy (DA-ST and DA-RN) vs. Standard deep learning models without domain adaptation (Swin-Transformer and ResNet) was evaluated on Atrial fibrillation detection accuracy on wearable ECG recordings. A domain-adversarial learning strategy based on the shifted window transformer (DA-ST) achieved an average atrial fibrillation detection accuracy of 98.67% on wearable ECG recordings, outperforming models without domain adaptation.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: