Registering Fiber Tracts with Fractional Anisotropy for Functional Sites Localization Matthieu Jomier 1 , Julien Jomier 2 , Guido Gerig 1 , and Stephen R. Aylward 2 1 Department of Psychiatry 2 Computer-Aided Diagnosis and Display Lab, Department of Radiology The University of North Carolina at Chapel Hill 27599 Chapel Hill, USA {matthieu,julien}@jomier.com gerig@cs.unc.edu, aylward@unc.edu Abstract. Localization of functional sites across individuals is impor- tant to assess and analyze neurological disparities in a population. Tradi- tionally this is accomplished by conducting anatomical registration and assuming functional to anatomical relationships. In this paper we in- vestigate the relationship between functional sites and fiber tracts. We propose a novel algorithm based on a model-to-image registration scheme to align diffusion tensor fiber tracts with fractional anisotropy map and thereby functional sites across individuals. We have evaluated our algo- rithm on ten normal controls, our results show similar accuracy compared to a global image-to-image registration strategy of fractional anisotropy images. 1 Introduction Diffusion Tensor Imaging (DTI) is a based on measurements of microscopic dif- fusion of water molecules that presumably indicate the direction of nerve bun- dles [8]. Registration of DTI is an important issue in medical imaging for inter- subject comparison or atlas creation. Two registration algorithms were proposed by Alexander et al. [12]. However, both methods assume that the spatial trans- formation is relatively small and the second method is sensitive to errors in a noisy environment. Others have used a fully elastic deformation field [13]-[15] for inter-subject registration of DT-MR images of different subjects. All three of these methods used the full DTI information to estimate spatial transformation of tensor fields. Fractional Anisotropy (FA) is one of the measurement most commonly used in clinical analysis of white matter diffusion properties. FA maps are generated from the baseline and six directional images after extraction of the diffusion tensor. FA maps are 3-dimensional images representing information about the shape of the diffusion tensor at each voxel with an intensity range between 0 and 1. In this paper, fractional anisotropy is used to compute a rigid plus scale transformation matrix (translation, rotation and scaling) for the registration of Submitted to MICCAI 2004 Not accepted for publication