Ultrasonic image processing using wavelet based deconvolution

The low resolution of the ultrasonic images is caused by the convolution between the reflect function of the measured tissues and the system pulse response. In this paper we estimate the system pulse response on the envelope ultrasonic image data using higher order spectra, and then implement the wavelet-based deconvolution algorithm. The resulting images behave clearer in velamen boundary and tissue layers, and the texture information also appears more detailed. The deconvolved image achieves good resolution gains in both axial and lateral direction.