Key result
The BayesBeat Bayesian deep learning model outperformed state-of-the-art methods in detecting atrial fibrillation from noisy photoplethysmography data, achieving an F1-score of 0.754 with an uncertainty threshold of 0.05.
Why the study?
Detecting AF in real time from smartwatch PPG data is challenged by inherent noise in the signals.
Does BayesBeat improve the accuracy of atrial fibrillation detection from noisy PPG signals compared to existing deep learning methods?
Does BayesBeat improve the accuracy of atrial fibrillation detection from noisy PPG signals compared to existing deep learning methods?
BayesBeat provides a highly efficient and robust Bayesian deep learning method for detecting atrial fibrillation from noisy smartwatch PPG data, outperforming current state-of-the-art models.
Answers are grounded in this paper and cite their sources.
May enable consumer-device AF screening; leaves open large-scale clinical validation before practice change.
Smartwatches or fitness trackers have garnered a lot of popularity as potential health tracking devices due to their affordable and longitudinal monitoring capabilities. To further widen their health tracking capabilities, in recent years researchers have started to look into the possibility of Atrial Fibrillation (AF) detection in real-time leveraging photoplethysmography (PPG) data, an inexpensive sensor widely available in almost all smartwatches. A significant challenge in AF detection from PPG signals comes from the inherent noise in the smartwatch PPG signals. In this paper, we propose a novel deep learning based approach, BayesBeat that leverages the power of Bayesian deep learning to accurately infer AF risks from noisy PPG signals, and at the same time provides an uncertainty estimate of the prediction. Extensive experiments on two publicly available dataset reveal that our proposed method BayesBeat outperforms the existing state-of-the-art methods. Moreover, BayesBeat is substantially more efficient having 40--200X fewer parameters than state-of-the-art baseline approaches making it suitable for deployment in resource constrained wearable devices.
No takes yet. Share an insight, caveat, or question.
Das et al. (2022) studied Atrial Fibrillation (n=175). BayesBeat (Bayesian deep learning model) vs. State-of-the-art deep learning models (DeepBeat, ResNeXt) was evaluated on F1-score for Atrial Fibrillation detection. The BayesBeat Bayesian deep learning model outperformed state-of-the-art methods in detecting atrial fibrillation from noisy photoplethysmography data, achieving an F1-score of 0.754 with an uncertainty threshold of 0.05.
Synapse has enriched 10 related papers. Consider them for comparative context: