ADAPTIVE PATTERN CORRELATION FOR 2-DIMENSIONAL BLOOD-FLOW MEASUREMENTS

One of the major drawbacks of ultrasonic Doppler instruments in measur ing blood flow is their inability to measure the velocity perpendicula r to the beam, Time domain RF echo or speckle tracking has been studie d as an alternative to overcome this problem. By acquiring two-dimensi onal (2-D) echo signals, both lateral (perpendicular to the beam) and axial (parallel to the beam) velocities can be calculated with 2-D pat tern correlation algorithms, One of the disadvantages of the current 2 -D pattern correlation algorithms is the extensive computation time in volved in computing the 2-D cross-correlation function, In this paper, we present several time-efficient bit-pattern correlation algorithms to execute 2-D speckle tracking. The proposed algorithms first estimat e the noise level from the acquired signals and use it as a priori kno wledge to minimize computation time. The reduction of computation time may make it more feasible for real-time measurements of how velocitie s in two dimensions, Radio frequency and video data collected from two commercial scanners are used to validate the feasibility of these pro posed algorithms with porcine blood as the flowing medium in in vitro experiments. The results obtained by the proposed algorithms are in go od agreement with those computed from the crosscorrelation function.