Tract-based spatial statistics of the olfactory brain in patients with multiple sclerosis.

Purpose: To investigate diffusion tensor abnormalities, e.g. fractional anisotropy (FA), mean diffusivity (MD), and radial diffusivity (RD), in olfactory structures of multiple sclerosis (MS) patients using diffusion tensor imaging (DTI).

Methods: Institutional review board-approved prospective study on 30 MS patients and 12 healthy controls investigated with MRI including DTI. Central olfactory structures were labelled on each patient's and healthy contro''s DTI volume. The diffusion tensor was determined in the central olfactory structures in MS patients. Tract-based spatial statistics (TBSS) was used to quantify the streamlines outgoing from the olfactory structures and to quantify changes in FA, MD, and RD within olfactory structures. These brain changes were correlated with olfactory function measured as TDI (Threshold, Discrimination, Identification) scores in patients and compared to our own reference group of 30 healthy volunteers.

Results: Central olfactory structures in the MNI (Montreal Neurological Institute) data volume comprise 4808 voxels (4808 mm(3)). TFCE (Threshold-free cluster enhancement) and cluster analysis of patients identified a total of 127 voxels in one cluster with a significantly decreased FA (p<0.05) and none for MD and RD within olfactory structures compared to healthy controls. The correlation with the age-normalised Identification subscore of the TDI score increased the significant number of voxels with decreased FA to 208 voxels, with increased MD to 370 and with increased RD 364 voxels at the same region.

Conclusion: The decrease in FA and increase of MD and RD correlate with the degree of identification impairment of olfactory function in MS patients and clusters of abnormalities were identified on a MNI data volume.