Why the study?
Can machine-learning algorithms effectively discriminate between high-quality and low-quality pulse waveforms obtained from wearable devices?
Population
Healthy young subjects
Comparison
Appropriate contact pressure (67.80 ± 1.55 mmHg) vs higher contact pressure (151.80 ± 3.19 mmHg)
Key result
Machine-learning analysis using a random-forest algorithm effectively discriminated between high-quality and low-quality pulse waveforms induced by varying contact pressures (AUC = 0.96).
Authors
Loading...
May enhance wearable pulse monitoring reliability; leaves open prospective outcome validation before clinical use.
Observational
Can machine-learning algorithms effectively discriminate between high-quality and low-quality pulse waveforms obtained from wearable devices?
Effect estimate: AUC 0.96
Machine-learning algorithms, particularly random forest, can accurately classify pulse waveform quality, which may improve the reliability of noninvasive physiological monitoring using wearable devices.
OuYoung et al. (2022) conducted an observational in Healthy. Machine-learning analysis was evaluated on Discrimination between high-quality and low-quality pulse waveforms (AUC 0.96). Machine-learning analysis using a random-forest algorithm effectively discriminated between high-quality and low-quality pulse waveforms induced by varying contact pressures (AUC = 0.96).