The identification, classification and size estimation of Doppler detected Microemboli

The aim of this thesis was to develop techniques to reliably identify, characterise and size emboli to help understand the clinical significance of such potentially devastating phenomena. Firstly a dual gated Doppler technique was developed to differentiate emboli from artefacts that produced excellent results when Doppler signals were analysed in the time domain. Several curious observations were noted, firstly the Doppler derived velocity did not correlate with the time of flight derived velocity, secondly the embolic signal duration in each sample gate was different, and finally an embolic signal was not always visible in both gates.;Another finding was the rapid frequency modulation apparent in some gaseous embolic signals. Several hypotheses have been proposed to explain this observation. Secondly attempts to differentiate gaseous from particulate emboli were investigated. The Wigner distribution analyser is presneted as an alternative to the conventional fast Fourier transform analyser and allows significant time resolution improvements. The sample volume length (SVL) of embolic signals was investigated as a new parameter for classifying emboli and when compared with previous parameter was found to produce the best sensitivity and specificity.;Finally the embolus to blood ratio (EBR) model is explored to assess its feasibility for sizing particulate emboli. Both dual and single frequency transducers were constructed and their beam characteristics were studied. Significant beam distortion and focusing occurred when skull bone was placed in front of the transducer which no longer produced uniform vessel insonation (a major assumption of the EBR model). In vitro experiments using the single frequency technique showed considerable EBR variability due to beam refraction and with the addition of intervening skull bone there were further errors introduced due to beam focusing effects. Statistical analysis showed that the experimental emboli could not be reliability separated, even into two size groups. Further work to improve the sizing of emboli is discussed.