k 3D Deformable Magnetic Resonance Textbook Based on Elasticity

complex shape variability evident in intersubject ̄ oanatomies can be accommodated using GrenanGlobal Shape models. Following the approach of :, 3, 4], we construct a digitized textbook containall of the information currently available about a :al neuroanatomy on a fixed 3D coordinate system. textbook contains anatomical information (mag: resonance images, cryosection blockfaces, etc...) symbolic information (structure names, functions, ̄ actions with other structures, etc...). The textanatomy is individualized by deforming the shape .1atomic atlas images to match corresponding irafrom an individual. Symbolic information in the ~ook is then deformed and projected onto the inlual’s anatomy using the same estimated trans.ation. In addition, functional information such ~sitron emission tomography (PET) images in the .~ individual can be mapped back to the textbook omy using the inverse transformation. By removhe individual shape variability from functional imthe relationship between function and anatomy ~s a normal population can be delineated. Once ocation of normal function is established, abhorties associated with disease states can be detected.