DETECTION AND MAPPING OF ABNORMAL BRAIN STRUCTURE WITH A PROBABILISTIC ATLAS OF CORTICAL SURFACES

Purpose: We have devised, implemented, and tested a technique for crea ting a comprehensive probabilistic atlas of the human cerebral cortex, based on high-dimensional fluid transformations. The goal of the atla s is to detect and quantify subtle and distributed patterns of deviati on from normal cortical anatomy, in a 3D brain image from any given su bject. Method: Given a 3D MR image of a new subject, a high-resolution surface representation of the cerebral cortex is automatically extrac ted. The algorithm then calculates a set of high-dimensional volumetri c maps, fluidly deforming this surface into structural correspondence with other cortical surfaces, selected one by one from an anatomic ima ge database. The family of volumetric warps so constructed encodes sta tistical properties of local anatomical variation across the cortical surface. Additional strategies are developed to fluidly deform the sul cal patterns of different subjects into structural correspondence. A p robability space of random transformations. based on the theory of ani sotropic Gaussian random fields, is then used to encode information on complex variations in gyral and sulcal topography from one individual to another. A complete system of 256(2) probability density functions is computed to reflect the observed variability in stereotaxic space of the points whose correspondences are found by the warping algorithm . Confidence limits in stereotaxic space are determined for cortical s urface points in the new subject's brain. Results: Color-coded probabi lity maps are generated, which highlight and quantify regional pattern s of deformity in the anatomy of new subjects. These maps indicate loc ally the probability of each anatomic point being as unusually situate d, given the distributions of corresponding points in the scans of nor mal subjects. 3D MRI Volumes are analyzed, from subjects with clinical ly determined Alzheimer disease and age-matched normal subjects. Concl usion: Applications of the random fluid-based probabilistic atlas incl ude the transfer of multisubject 3D functional, vascular, and histolog ic maps onto a single anatomic template, the mapping of 3D atlases ont o the scans of new subjects, and the rapid detection, quantification, and mapping of local shape changes in 3D medical images in disease and during normal or abnormal growth and development.