In this paper a modular approach of segmentation which combines the Bayesian model with the deformable model is proposed.
It is based on the level set method, and breaks up into two great parts. Initially, a preliminary stage allows constructing
the information map. Then, a deformable model, implemented with the Generalized Fast Marching Method (GFMM), evolves towards
the structure to be
... [Show full abstract] segmented, under the action of a force defined from the information map. This last is constructed from
the posterior probability information. The major contribution of this work is the use and the improvement of the GFMM for
the segmentation of 3D images and also the design of a robust evolution model based on adaptive parameters depending on the
image. Experimental evaluation of our segmentation approach on several MRI volumes shows satisfactory results.