Key result
A 50-layer convolutional neural network accurately detected atrial fibrillation episodes from continuous ambulatory photoplethysmography signals, achieving a test AUC of 94.8% and demonstrating robustness to motion artifacts.
Why the study?
Does a deep learning model applied to wearable PPG signals improve the detection of atrial fibrillation compared to a feature-based baseline in ambulatory free-living conditions?
Observational (n=81)
Does a deep learning model applied to wearable PPG signals improve the detection of atrial fibrillation compared to a feature-based baseline in ambulatory free-living conditions?
A deep learning algorithm applied to continuous wrist-worn PPG signals can accurately detect atrial fibrillation in ambulatory settings with high robustness to motion artifacts.
Answers are grounded in this paper and cite their sources.
Enables potential continuous AF monitoring via wearables; leaves open prospective clinical validation.
We develop an algorithm that accurately detects Atrial Fibrillation (AF) episodes from photoplethysmograms (PPG) recorded in ambulatory free-living conditions. We collect and annotate a dataset containing more than 4000 hours of PPG recorded from a wrist-worn device. Using a 50-layer convolutional neural network, we achieve a test AUC of 95% and show robustness to motion artifacts inherent to PPG signals. Continuous and accurate detection of AF from PPG has the potential to transform consumer wearable devices into clinically useful medical monitoring tools.
No takes yet. Share an insight, caveat, or question.
Voisin et al. (2018) conducted an observational in Atrial Fibrillation (n=81). Deep learning algorithm (50-layer CNN) for PPG vs. Feature-based baseline algorithm (IBI RMSSD) was evaluated on Detection of Atrial Fibrillation (AUC on test set) (AUC 94.8%). A 50-layer convolutional neural network accurately detected atrial fibrillation episodes from continuous ambulatory photoplethysmography signals, achieving a test AUC of 94.8% and demonstrating robustness to motion artifacts.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: