This thesis is concerned with an investigation of the effect of ultrasonic pulse bandwidth on the human blood velocity measurement in an ultrasonic Doppler system. While the Doppler shifted frequency in a continuous wave Doppler system can be represented in terms of only the center frequency of an ultrasonic piezoelectric transducer, that of a pulsed Doppler system comprises a number of various single frequency components around the center frequency because of its inherent nature of a burst pulse, rather than CW, excitation. The above statement is analytically formulated for measuring blood velocities due to a cloud of random scatterers having either a single uniform velocity or a distribution of diverse velocities across the depth of interest in a human blood vessel.
We propose the second order sampling technique, which is amenable to hardware implementation, for evaluating such various parameters as mean, standard deviation, and sideband separation index of Doppler shifted spectrum and compare its performance with that of the conventional quadrature detection method which uses a mixing scheme. The proposed second order sampling technique as applied to a Doppler front-end system for velocity measurement is verified through computer simulation on a VAX-11/750 computer.
When a wideband pulse is used to interrogate the blood vessel, the second order sampling technique is inferior in velocity estimation capability to the quadrature detection method, but for narrowband ultrasonic pulses employed the former is almost comparable to the latter.
