MRI-based volumetry of structural MR images can reveal the pattern of atrophic changes in patients with atypical parkinsonian syndromes. Using a region-of-interest (ROI) analysis, it has been shown that the patterns of atrophy in predefined regions of interest allow the separation PD from MSA and PSP (36) as well as PSP from CBD (37). For instance, a recent study by Groschel and colleagues demonstrated that the volumes of midbrain, parietal white matter, temporal gray matter, brainstem, frontal white matter, and pons separate best between patients with PSP or CBD (37). A stepwise linear discriminant analysis resulted in two canonical discriminant functions, which allowed for the correct prediction of the diagnosis in 95% of healthy control subjects as well as in 76% of all PSP and 83% of all CBD patients (37). The discriminant functions revealed similar results in patients with definite PSP/CBD and in patients with possible and probable PSP/CBD (37). Prospective studies are now needed to show whether MRI-based volumetry will have clinical applicability.
In recent years, voxel-based morphometry (VBM) of structural MRI data has been introduced as a simple and objective approach for characterizing small-scale differences in white and gray matter (36). VBM refers to a voxel-wise statistical comparison of the local concentration of gray (or white) matter between two groups of subjects. The procedure involves spatial normalization of high-resolution MR images into the same stereotactic space followed by segmentation of the gray (or white) matter. The smoothed gray matter segments are then compared using parametric statistical tests and the theory of Gaussian random fields (38). Other voxel-based morphometric approaches, such as deformation-based or tensor-based morphometry, allow the study of regional differences in brain shapes between groups of subjects (38). In contrast to conventional ROI analyses, VBM and related approaches provide whole-brain coverage and are highly sensitive to subtle structural changes within a single brain region. Since VBM is an automated procedure, the analysis is highly observer independent. Finally, VBM is not affected by partial volume effects. VBM cannot only be applied to highresolution structural MRI, but VBM can also be used to investigate between-group differences of ADC maps or fractional anisotropy maps. This allows to screen for regional changes in neuronal integrity (i.e., VBM of ADC maps) or regional changes in fiber orientation (i.e., VBM of DT-weighted MRI).
Was this article helpful?