| Summary: | Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2017. === Cataloged from PDF version of thesis. === Includes bibliographical references (pages 155-163). === While blood pressure is commonly used by doctors as an indicator of patient health, the available techniques to measure the quantity suffer from many inconveniences such as cutting off blood flow, being cumbersome to use, being invasive, or being inaccurate. The research addresses many of these inconveniences by developing and evaluating a novel ultrasound-based blood pressure measurement technique that is non-invasive and non-occlusive. The technique proceeds in three steps: data acquisition, data reduction, and optimization. In the data acquisition step, an ultrasound probe is placed on a patient's artery and a force sweep is conducted such that the contact force gradually increases; both the applied force and B-Mode images are recorded. In the data-reduction step, the Star-Kalman filter is applied in order to find the size of the artery in each image frame captured. The segmentation data and contact force data are inputs into the optimization step which consists of two sequential optimizations; the first makes many modeling assumptions and gives an estimate of pulse pressure while the second makes less assumptions and uses the approximation of pulse pressure to obtain absolute values of systolic and diastolic blood pressure. Central to the optimization algorithm is a computational biomechanical model of the artery and surrounding tissue, which is numerically modeled using finite elements. The impact of major modeling assumptions is corrected with a one time calibration. The technique is validated on a number of different data sets. Major data sets discussed include data taken on the carotid artery of (1) 24 single-visit nominally healthy volunteers, (2) two multi-visit nominally healthy volunteers, (3) one multi-visit hypertensive volunteer, and (4) one multi-visit hypotensive volunteer; additional miscellaneous data sets are taken and analyzed as part of this dissertation. The algorithm performance is quantified against readings from an automatic oscillometric cuff. Results show that systolic and diastolic blood pressures can be predicted by the algorithm. The technology discussed in this dissertation represents a proof-of-concept of a blood pressure measurement technique that could occupy a clinical middle ground between the invasive catheter and cuff-based techniques. === by Aaron Michael Zakrzewski. === Ph. D.
|
|---|
