Mapping image data to stereotaxic spaces: applications to brain mapping.

A methodology for spatial normalization of image data is presented. This methodology is based on a map between homologous features of an individual brain and the target brain, which is used to drive a three-dimensional elastic warping transformation. Functional or structural information present in the original, nonnormalized images is preserved during this transformation. In particular, information such as the volume or the total amount of a radioactive agent in any brain region can be calculated directly from the normalized images. Moreover, subtle morphological characteristics of an individual brain are captured by the properties of the spatial transformation applied to that brain. Intersubject or interpopulation comparisons are performed by comparing the corresponding transformations.