Model creation and deformation for the automatic segmentation of the brainin MR images

In this paper a method for the automatic segmentation of the brain in magne tic resonance images is presented and validated. The proposed method invoic es two steps 1) the creation of an initial model and 2) the deformation of this model to fit the exact contours of the brain in the images. A new meth od to create the initial model has been developed and compared to a more tr aditional approach in which initial models are created by means of brain at lases, A comprehensive validation of the complete segmentation method has b een conducted on a series of three-dimensional T1-weighted magnetization-pr epared rapid gradient echo image volumes acquired both from control volunte ers and patients suffering from Gushing's disease. This validation study co mpares results obtained with the method we propose and contours drawn manua lly. Averages differences between manual and automatic segmentation with th e model creation method we propose are 1.7% and 2.7% for the control volunt eers and the Gushing's patients, respectively, These numbers are 1.8% and 5 .6% when the atlas-based method is used.