A method for segmenting magnetic resonance (MR) images of an anatomical body of interest, for example the brain, comprises classifying a plurality of selected structures within the body of interest based on a plurality of image processing computations wherein the computations are adapted to relate respective T2 relaxation times corresponding to each of the structures. Thereafter, the method comprises segmenting the MR images for each of the structures substantially concurrently based on the plurality of image computations. A system for automatically segmenting magnetic resonance (MR) images of an anatomical body of interest comprises a processor coupled to an MR image acquisition device wherein the processor is adapted to perform concurrent segmentation computations for a plurality of selected structures within the anatomical body of interest. The system further comprises an interface unit coupled to the processor adapted to present information relating to the segmented computations corresponding to the plurality of selected structures.