The purpose of the study was to asses the potential use of pulse wave velocity (PWV) and digital volume pulse (DVP) as estimators of systolic (SBP) and diastolic (DPB) blood pressure. Single and multiple correlation studies were conducted, including biometric parameters and risk factors. Brachial-ankle PWV (baPWV) and DVP signals were obtained from a Pulse Trace PWV and Pulse Trace PCA (pulse contour analysis), respectively. The DVP (obtained by photoplethysmography), allowed stiffness (SI) and reflection indexes (RI) to be derived. The first study on 47 healthy volunteers showed that both SBP and DPB correlated significantly both with baPWV and SI. Multiple regression models of the baPWV and the waist-to-hip ratio (WHR) allowed SBP and DBP to be modeled with r = 0.838 and r = 0.673, respectively. SI results also employed WHR and modeled SBP and DBP with r = 0.852 and r = 0.663, respectively. RI did not correlate either with SBP or DBP. In order to avoid the use of ultrasound techniques to measure PWV, we then developed a custom-built system to measure PWV by photoplethysmography and validated it against the Pulse Trace. With the same equipment we conducted a second pilot study with ten healthy volunteers. The best SBP multiple regression model for SBP achieved r = 0.997 by considering the heart-finger PWV (hfPWV measured between R-wave and index finger), WHR and heart rate. Only WHR was significant in the DBP model. Our findings suggest that the hfPWV photoplethysmography signal could be a reliable estimator of approximate SBP and could be used, for example, to monitor cardiac patients during physical exercise sessions in cardiac rehabilitation.

A computer based photoplethysmographic analyzer was developed. The signal was obtained by infrared light through the finger. It was converted into digital domain by a signal processing circuitry. First, second and fourth derivatives of the signal were computed. The acquisition and mathematical processing programs were implemented in object-oriented programming C. A study with 38 people, 19 healthy volunteers and 19 subjects with previously diagnosed cardiovascular disease (atherosclerosis, hypertension, diabetes mellitus) was carried out. A t-tested distribution between healthy volunteers and patients showed a significant differences in calculated parameters. Photoplethysmographic augmentation index has shown to be a noninvasive parameter for vascular assessments. Using derivatives, the inflection points identification of the digital volume pulse is simpler and more accurate.

No abstract

Abstract