Key result
ResNet18 2D deep learning assesses PPG signal quality with ~99% accuracy, outperforming traditional SVM.
Why the study?
Movement artifacts in PPG signals can cause false interpretations or unnecessary signal discarding, which hinders the detection of episodic arrhythmias such as paroxysmal AF.
Does a 2D image-based deep learning model (ResNet18) improve the accuracy of PPG signal quality assessment in the presence of atrial fibrillation compared to 1D time-series machine learning (SVM)?
Does a 2D image-based deep learning model (ResNet18) improve the accuracy of PPG signal quality assessment in the presence of atrial fibrillation compared to 1D time-series machine learning (SVM)?
p-value: p=0.0117
Translating time-series PPG data into 2D images for deep learning algorithms (ResNet18) yields highly accurate signal quality assessments, which may enhance reliable atrial fibrillation detection by appropriately filtering movement artifacts.
2D DL may enhance PPG quality filtering in AF; hypothesis-generating and requires prospective validation before clinical use.
OBJECTIVE: Photoplethysmography (PPG) monitoring has been implemented in many portable and wearable devices we use daily for health and fitness tracking. Its simplicity and cost-effectiveness has enabled a variety of biomedical applications, such as continuous long-term monitoring of heart arrhythmias, fitness, and sleep tracking, and hydration monitoring. One major issue that can hinder PPG-based applications is movement artifacts, which can lead to false interpretations. In many implementations, noisy PPG signals are discarded. Misinterpreted or discarded PPG signals pose a problem in applications where the goal is to increase the yield of detecting physiological events, such as in the case of paroxysmal atrial fibrillation (AF)-a common episodic heart arrhythmia and a leading risk factor for stroke. In this work, we compared a traditional machine learning and deep learning approaches for PPG quality assessment in the presence of AF, in order to find the most robust method for PPG quality assessment. APPROACH: The training data set was composed of 78 278 30 s long PPG recordings from 3764 patients using bedside patient monitors. Two different representations of PPG signals were employed-a time-series based (1D) one and an image-based (2D) one. Trained models were tested on an independent set of 2683 30 s PPG signals from 13 stroke patients. MAIN RESULTS: ResNet18 showed a higher performance (0.985 accuracy, 0.979 specificity, and 0.988 sensitivity) than SVM and other deep learning approaches. 2D-based models were generally more accurate than 1D-based models. SIGNIFICANCE: 2D representation of PPG signal enhances the accuracy of PPG signal quality assessment.
No takes yet. Share an insight, caveat, or question.
Pereira et al. (2019) studied Atrial Fibrillation (n=3,777). 2D image-based Deep Learning (ResNet18) vs. 1D time-series traditional Machine Learning (SVM) was evaluated on Accuracy of PPG signal quality assessment (p=0.0117). The ResNet18 deep learning model utilizing 2D image representations achieved 98.5% accuracy for PPG signal quality assessment, significantly outperforming traditional SVM approaches (p=0.0117).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: