Key result
Gaussian process regression using PPG features accurately estimates RR and SpO2.
Why the study?
Continuous monitoring of respiratory rate and oxygen saturation is crucial for patients with cardiac, pulmonary, and surgical conditions, and photoplethysmogram signals have been recommended for their evaluation.
Can machine learning models accurately estimate respiratory rate and blood oxygen saturation from photoplethysmogram signals?
Comparison
19 machine learning models trained using PPG signal features
Design
Machine learning model development and comparison study
Authors
Loading...
PPG-based models may enable low-cost continuous monitoring; leaves open prospective clinical validation before adoption.
Can machine learning models accurately estimate respiratory rate and blood oxygen saturation from photoplethysmogram signals?
A Gaussian process regression model can accurately estimate respiratory rate and blood oxygen saturation from PPG signals, potentially offering a low-cost, non-invasive monitoring solution.
Shuzan et al. (2023) studied this question. Gaussian process regression model using photoplethysmogram (PPG) signal features vs. Other machine learning models was evaluated on Respiration rate (RR) and blood oxygen saturation (SpO2) estimation accuracy. A Gaussian process regression model using photoplethysmogram signal features estimated respiration rate with a mean absolute error of 0.89 and blood oxygen saturation with a mean absolute error of 0.57.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: