Vývoj metodiky kontinuálního neinvazivního měření krevního tlaku

Abstract

The main objective of this work is to find a new methodology for measuring continuous non-invasive blood pressure based on the pulse wave velocity in the vascular system. The work is based on the literature research of the basic model for the determination of non-invasive continuous blood pressure based on the measurement of pulse transit time. From the information obtained from the review, the methodology of measuring the pulse transit time/pulse wave velocity was modified in order to achieve more accurate results and to reduce the human factor that causes significant inaccuracy due to imperfect sensor placement. The review discusses in detail the models for continuous non-invasive blood pressure estimation and their modifications to ensure increased accuracy. In particular, model modifications include input parameters describing blood circulation - systemic vascular resistance, vascular elasticity, and vascular stiffness. The thesis deals with modifications to the existing physical vascular model to more closely mimic the real vascular system of the human body. These modifications include the baroreflex function or the simulation of different wall hardness of artificial arterial segments. As this is a simulation model of the vascular system, the measurement of pressure and volume pulse wave is also an important step, where it is not possible to use photoplethysmography method due to the absence of light absorbing particles. Based on the experimental measurements for different settings of the vascular model, pulse wave measurements were performed using pressure and capacitive sensors with subsequent processing of the measured signals and detection of the pulse wave features. Predictive regression models were established based on the pulse wave features and showed sufficient accuracy in their determination, followed by two methods for obtaining the parameter on the hardness of the vascular wall based on the measurable parameters. The first method was a predictive regression model, which showed an accuracy of 74.1 %, and the second method was an adaptive neuro-fuzzy inference system, which showed an accuracy of 98.7 %. These pulse wave velocity determinations were verified by further direct pulse wave measurements and the results were compared. The dissertation results in the determination of pulse wave propagation velocity using only one plethysmographic sensor without the need for measurements at two different locations with accurate distance measurements and the possibility of application in clinical practice.