A Statistical Model to Estimate Flow Mediated Dilation Using Recorded Finger Photoplethysmogram

摘要

Noninvasive measurement of brachial artery (BA) flow mediated dilation (FMD), commonly measured by high resolution ultrasonography, has provided a non-invasive means for the evaluation of endothelial function. Recently, another technique based on fmger photopletysmogram (PPG) analysis has been proposed by the authors. In this paper, simultaneous measurement of BA FMD and PPG signals data from 86 human subjects (43 healthy and 43 risk) is analyzed and a statistical model is proposed to estimate the BA FMD based on PPG signals, the BA diameter at baseline as well as other demographics factors such as age and gender. Linear regression by multivariate analysis was conducted to obtain the variables that were significantly related to BA FMD. The mean of absolute error for the model was 0.16 + 0.12 and the estimated data (via PPG) is correlated to the measured one (via ultrasound) with a correlation factor R = 0.725. As a confirmation, it was found that the estimated FMD for the risk group (22.3 + 3.4 %) was significantly lower than that of the healthy group (25.9 + 3.5 %),p < 0.01, whereas a similar trend was observed from the measured data: risk group (22.3 + 5.2 %) versus healthy group (26.1 + 7.8 %),p < 0.001. The model can estimate peak BA (ultrasound) FMD from the information of the baseline BA diameter and peak PPG AC, allowing for an alternative means of evaluating the endothelial function.