Investigation on Pulse Wave Forward Peak Detection and Its Applications in Cardiovascular Health.

Abstract

The contours of the pulse wave vary greatly, which affect the accuracy of pulse wave peak detection and the reliability of subsequent peak-based cardiovascular health analyses. We proposed an algorithm to reliably detect the peak of forward pulse wave (forward peak) and proposed to use it for improving the accuracy in cardiovascular health analysis. A method based on Gaussian fitting was proposed to detect the forward peak. Then, the forward peak was utilized for instantaneous heart rate (HR), heart rate variability (HRV), and augmentation index (a cardiovascular risk marker reflecting arterial stiffness) estimations. The accuracy of HR/HRV obtained by forward peak was compared with that obtained by other photoplethymogram (PPG) characteristic points previously reported, using electrocardiogram-derived HR/HRV as gold standard. The correlation between augmentation index and age was calculated. The performance was verified using PPG-based pulse wave data with different contours while they were recorded at different locations from subjects with a wide range of age. The proposed forward peak detection method had smaller estimation error when compared with the gold standard, than other PPG characteristic points in estimating HR/HRV. The augmentation index extracted from the proposed forward peak was significantly correlated with age (p < 0.01). The proposed algorithm can relatively reliably detect the forward peak and has a wide application prospect in cardiovascular health. Due to the convenience of PPG measurements, this proposed forward peak detection method has the potential to be widely used in the fields of wearable devices and telemedicine.