Deconvolution of in vivo ultrasound images

In an ultrasound image, the influence of the pulse and attenuation should be removed from the picture in order to display a more consistent and uniform image. The author describes an algorithm to remove the influence of the attenuated pulse on the image. The algorithm takes into account the varying pulse, noise in the acquired signal, and the changing reflectivity in the tissue. Both one- and two-dimensional processing can be implemented. The algorithm relies on prior knowledge of the pulse and of the covariance of the noise and the reflections. Algorithms to estimate these factors are given. Examples of 1-D deconvoluted pictures of phantom data and in vivo data are given. They show, especially for the phantom data, an increased contrast and resolution