Doppler Ultrasound Evaluation of the Haemodynamics of Fainting
Degree GrantorUniversity of Canterbury
Degree NameMaster of Science
Every time we stand up, gravity pulls nearly a quarter of the bodies blood supply into the lower body. The heart and circulation must respond within seconds to keep enough blood flowing up to the brain. To meet the challenge the heart will speed up to 10 – 15 beats a minute and the nervous system cause the arteries to narrow so that blood pressure rises. If there is an abnormal response due to a temporary malfunction of the nerves and arteries, blood pressure may drop and fainting can briefly occur. Tilt testing is often performed to try to reproduce an episode of fainting in patients with this abnormal response. Blood pressure, heart rate, nerve activity and measures of the Superior Mesenteric Artery (SMA) Doppler waveform can be monitored with the subject lying horizontal, and tilted to 60o to try to trigger an episode of loss of consciousness. The Doppler arterial waveform images can be analysed to provide information about blood flow in or out of the gut (a large vascular bed) and related, along with the other physiological measurements, to the fainting process. A simple ratio, such as end diastolic velocity to peak systolic velocity could be used to measure changes of SMA blood flow. However this simple measure may not be sufficiently discriminatory and a more sophisticated measure, such as the ratio of the area under the curve in systole versus that in diastole will hopefully describe more completely the changes taking place in diastole and hence give a measure of vasoconstriction or vasodilation of the SMA. In order to do this the Doppler waveforms obtained over time from the SMA need to be processed off-line after being saved as TIFF files for example. Then image processing on the waveform as it appears on the TIFF file, using Matlab, extracts the maximum frequency envelope. From this envelope the cut-off between systole and diastole is determined, the areas of interest calculated and hence the above ratio. The challenge is to extract a maximum velocity envelope for a range of images, some ‘noisy’, by interactively selecting the better cycles in any given image rather than completely automating the process. The purpose of the research reported here is to create or modify image enhancement techniques to allow a better evaluation of the Doppler waveforms. Three or four methods of producing a maximum velocity envelope will be compared and one chosen on the basis of its speed and ability to accurately process noisy images. The area ratio calculated will also be investigated over a number of images to investigate its sensitivity and measure of blood flow changes.