Estimation of complex ultrasonic medium responses by deconvolution

The resolution of a conventional ultrasound system is limited by the pulse length and beam width, and significant improvement is possible by deconvolution of the ultrasound RF signal. Most previous research has been based on the assumption that the medium response was real valued. However, a complex-valued medium response incorporates the phase of the carrier wave and provides a more comprehensive and appropriate model even though the deconvolution becomes substantially more difficult because the problem is inherently underdetermined. To address this problem, an iterative algorithm, the least squares method with point count regularization (LSPC) was developed and presented. Simulations and experiments were used to examine the performance of the algorithm, and excellent results were obtained and reported. Further study of the algorithm to reduce computational complexity and to improve robustness is in progress.