The Effect of Local Fiber Model On Population Studies
Malcolm, James G.
Shenton, Martha E.
MetadataShow full item record
Diffusion tensor imaging has made it possible to evaluate the organization and coherence of white matter fiber tracts. Hence, it has been used in many population studies, most notably, to find abnormalities in schizophrenia. To date, most population studies analyzing fiber tracts have used a single tensor as the local fiber model. While robust, this model is known to be a poor fit in regions of crossing or branching pathways. Nevertheless, the effect of using better alternative models on population studies has not been studied. The goal of this paper is to compare white matter abnormalities as revealed by two-tensor and singletensor models. To this end, we compare three different regions of the brain from two populations: schizophrenics and normal controls. Preliminary results demonstrate that regions with significant statistical difference indicated using one-tensor model do not necessarily match those using the two-tensor model and vice-versa. We demonstrate this effect using various tensor measures.