Segmentation of fetal 3D ultrasound based on statistical prior and deformable model

A statistical variational framework is proposed for the fetus and uterus segmentation in ultrasound images. The Rayleigh and exponential distributions are used to model the pixel intensity. An energy is derived to perform an optimal partition of the 3D data into two classes corresponding to these two distributions, in a Bayesian MAP framework. Some numerical difficulties are raised by the combination of heterogeneous distributions in a variational level-set formulation, as discussed in the paper. Results on simulated and real data are presented and show that assuming different distributions provides better results than with the sole Rayleigh distribution.