Date of Award

Spring 2019

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical & Aerospace Engineering

Program/Concentration

Mechanical Engineering

Committee Director

Julie Zhili Hao

Committee Member

Leryn Reynolds

Committee Member

Linda Vahala

Committee Member

Thomas Alberts

Abstract

This dissertation presents a model-based method for estimating arterial wall parameters from noninvasively measured arterial pulse signals via a microfluidic-based tactile sensor. The sensor entails a polydimethylsiloxane (PDMS) microstructure embedded with 5×1 transducer array built on Pyrex/Polyethylene terephthalate (PET) substrate. The arterial pulse causes a time-varying deflection on the top of the PDMS microstructure, which registers as a resistance change by the transducer at the site of the artery.

Owing to the time-harmonic nature of its radial motion, the arterial wall is modeled as a second-order dynamic system. By combining this dynamic model with a hemodynamic model of blood flow, the arterial wall parameters: elasticity, viscosity, and radius, were represented by spring constant and damping coefficient. Moderate exercise was utilized to introduce changes in arterial wall parameters because sensor-artery interaction varies with arteries and subject-specificity (i.e. overlying tissue and blood pressure). The same sensor was used to measure pulse signals pre-exercise and post-exercise on a subject for tracking changes of arterial wall parameters. The estimated values on five subjects revealed a statistically significant difference between pre-exercise and post-exercise (p<0.05).

A measured arterial pulse signal is a combination of the sensor design, hold-down pressure, and subject-specificity. Optimization of the sensor design was needed to improve the sensor-artery interaction for maximizing the measured pulse amplitude. Toward this end, the same sensor design with a PDMS conformal layer of different mixing ratios (elastomer curing agent and base) and different thickness was tried to improve the sensor-artery interaction in six subjects with different BMI. Adding a conformal layer was found to significantly increase the amplitude of a measured pulse signal in a subject with high BMI. The changes in arterial wall parameters introduced by moderate exercise were estimated from the measured signals using the sensor with a conformal layer on two subjects. The obtained changes in arterial wall parameters were consistent with the measured changes using a medical instrument.

In summary, the feasibility of a model-based method for estimating arterial wall parameters from measured arterial pulse signals was validated using the changes introduced by moderate exercise on several subjects. Future work will focus on implementing the sensor for practical use.

DOI

10.25777/8y1e-hm65

Available for download on Friday, July 02, 2021

Share

COinS