An investigation of the Doppler signal power method for detecting changes in the size of the middle cerebral artery

The Transcranial Doppler ultrasound technique is widely used for detecting velocity changes in the middle cerebral artery. However, flow changes inferred from velocity alone may be subject to error if the cross-sectional area of the vessel changes during a recording. One potentially ideal method for measuring changes in middle cerebral artery size is the Doppler signal power method. This is based on the theory that the power of a Doppler signal is proportional to the volume of blood from which the signal originates. In practice a number of factors other than vessel size influence signal power and hence may invalidate the technique. This thesis documents an investigation of the feasibility of using the Doppler signal power to measure changes in middle cerebral artery size. The main factors influencing signal power were considered. In-vitro recordings from a wall-less flow phantom showed a non-proportional relationship between power and channel size. This was deduced to be caused primarily by non-uniform insonation and high pass filtering. Following from these results, non-uniform insonation was identified as the main source of error for in-vivo recordings. An investigation of the effects of temporal bone on beam shape showed that beam shape across the middle cerebral artery is likely to be highly distorted and vary unpredictably between individuals. Theoretical modelling was used to predict the errors caused by beam shape in power changes detected from the middle cerebral artery, and demonstrated a variable error magnitude dependent on beam shape, vessel size, vessel position and beam angle. A novel technique for correcting the Doppler power spectrum for the effects of beam shape was proposed, and performed well when tested with theoretical and in-vitro spectra. Finally, an initial investigation of the correction technique using in-vivo spectra provided important information regarding future investigations of the in-vivo use of the Doppler power method.