ML estimation for acoustical de-blurring

In this paper we present a deconvolution technique for ultrasound images based on a maximum likelihood estimation procedure. In our approach the blur effect is estimated through homomorphic techniques either from the transducer response, measured with an experimental setting, or from the in-vivo scans. The ultrasonic signal envelope is then considered as a discrete sequence affected by a known intersymbol interference and additive white Gaussian noise and processed with a reduced complexity Viterbi algorithm, which is an optimum solution for the estimation of discrete sequences affected by both noise and intersymbol interference.