RESUMO
Given the gap between the crucial role of measuring arterial stiffness in cardiovascular disease prevention and the lack of a technology for frequent/continuous measurement to assess it without an operator, we have developed a wearable accelerometer-based system. It estimates local stiffness metrics (Ep, ß, and AC) by employing a one-point patient-specific calibration on the features of acceleration plethysmogram (APG) signal. An in-vivo study on 12 subjects was conducted (a) to select suitable ones from the host features on which the calibration could be applied and (b) to assess the feasibility of reliably estimating the stiffness metrics post-exercise when calibrated prior. The acquired APG signals were found to be reliable (SNR > 38 dB) and repeatable (CoV < 10 %). By examining a correlation matrix, it was found that (a-b)/(a"-b") is a potential feature of consideration for calibration against the stiffness. Due to exercise intervention, the local stiffness metrics have physiologically perturbed by a significant amount (p < 0.05), as observed from the reference measurements. Estimated Ep was found to have statistically significant and strong correlation (r = 0.761, p < 0.05) with actual Ep value, whereas statistically significant and moderate correlation were found with estimated ß (r = 0.682, p < 0.05) and estimated AC (r = 0.615, p < 0.05) with their respective actual measures. The system demonstrated its ability to estimate post-exercise stiffness metrics using the baseline calibration, even when subject to significant physiological changes.Clinical Relevance- This study reveals the potential of the developed wearable system to be used for continuous stiffness estimation even in the presence of hemodynamic perturbations.