Analysis of the pulse transmit time

OData support
Supervisor:
Dr. Jobbágy Ákos Andor
Department of Measurement and Information Systems

This BSc thesis assesses how the pulse transit time can be used to characterise the cardiovascular system of a person. I have used the ECG and PPG signals of the examined person that have been recorded simultaneously in order to determine the pulse transmit time. To achieve this, my goal was to appoint the appearance of the R wave in the ECG signal and the arrival of the pulse wave to the fingertip.

During my work it was necessary to evaluate the recordings made by the HHMD (Home Health Monitoring Device) of the Medical Technology Laboratory at the Department of Measurement and Information Systems. It was also necessary to expand the existing database and to make new recordings.

Most of the recordings evaluated by me contain ECG and PPG signals that were recorded during blood pressure measurement. It means that a cuff wrapped around the examined patient’s upper arm was slowly inflated and deflated.

In the course of my analysis, I compared recordings of patients affected by cardiovascular disease, healthy seniors and healthy young persons, using MATLAB. I processed recordings of 6 patients affected by cardiovascular disease, 5 healthy seniors and 17 healthy young persons. Approximately

1500 recordings have been processed. As the first step of the analysis I selected roughly 170 recordings with good enough signal-to-noise ratio for algorithm development.

During my work I have defined several parameters in order to characterize the examined patients based on the measurement results. I have particularly focused on the differentiation between patients and healthy persons based on the recordings. Most of the defined parameters did not prove to be applicable for the task. However, in many cases these parameters have shown results contrary to our previous experiences and they have brought up several questions requiring further investigation.

The best parameter is the one characterizing the change in the slope of the cuff pressure – pulse transit time plot.

I developed an algorithm to estimate the diastolic blood pressure based on the change of pulse transmit time. In its present form the accuracy of the estimation is poor, further development is needed.

In my future work, my aim is to improve the quality of recordings made by the HHMD, by means of the changing of its compression algorithm. During my thesis work several new research tasks have been defined. I intend to take part in the related research work.

Downloads

Please sign in to download the files of this thesis.