Table 1.
Demographic and clinical data of subjects.
Fig 1.
Graphical representations of the tracts studied.
Sagittal (upper) and coronal (lower) view. The uncinate fasciculus (green), the anterior cingulum (orange), the inferior frontooccipital fasciculus (light blue) and the forceps minor (dark blue). Tracts are from TrackVis, overlaid on high resolution images for illustrative purposes.
Fig 2.
Cortical thickness and Hayling error score correlations.
Correlations between cortical thickness and total error score on the Hayling test in patients with behavioural variant frontotemporal dementia and progressive supranuclear palsy using Freesurfer. Correction for multiple comparisons was made using the Monte Carlo method at the cluster level, at p < 0.01 (z-vertex 1.3). Age was entered as a nuisance variable. Coloured areas represent significant negative correlations, with the scale bar representing p on a logarithmic scale. Only the right hemisphere is shown, from lateral, medial and inferior views. No regions with significant correlation with cortical thickness were detected in the left hemisphere.
Fig 3.
Fractional anisotropy and Hayling error score correlations.
Scatter plots of total error score on the Hayling test and fractional anisotropy (FA) of tracts in patients with behavioural variant frontotemporal dementia (dots) and progressive supranuclear palsy (triangles), for A: the right uncinate fasciculus, B: the right anterior cingulum, and C: the forceps minor. B (adjusted) and p are derived from the linear regression model, with age as covariate.
Fig 4.
Frontal Behavioural Inventory and FA of tracts.
Scatter plots of Frontal Behavioural Inventory composite score of items 12–22 (FBI12-22) and fractional anisotropy (FA) of tracts in patients with behavioural variant frontotemporal dementia (bvFTD) (dots) and progressive supranuclear palsy (PSP) (triangles), for A: the right uncinate fasciculus, B: the left uncinate fasciculus, C: the anterior cingulum and D: the forceps minor. B (adjusted) and p are derived from the linear regression model, with age as covariate.
Fig 5.
Cortical thinning in bvFTD patients.
Results of the comparison of cortical thickness between patients with behavioural variant frontotemporal dementia and healthy controls. Results are from the Freesurfer analysis, results of the general linear model analysis were corrected for multiple comparisons at the cluster level using the Monte Carlo method for p-cluster at p < 0.01 (z-vertex 2.0). No nuisance variables were entered into the model. Coloured areas represent areas of significant differences, where warmer colours represent cortical thickening and cooler colours cortical thinning. Scale bar represents p on a logarithmic scale. The upper row is the right hemisphere, the lower row is left hemisphere, from lateral, medial and inferior views.
Fig 6.
Boxplots of FA values of tracts studied.
HC: healthy controls, PSP: progressive supranuclear palsy, bvFTD: behavioural variant frontotemporal dementia. FA: fractional anisotropy. FM: forceps minor, UF: uncinate fasciculus, IFOF: inferior frontooccipital fasciculus, aCi: anterior cingulum. Rh: right hemisphere and lh: left hemisphere. Boxes represent 25th and 75th percentile with median, whiskers minimum and maximum value. Staples represent statistically significant differences in between group pairs, at p < 0.05, uncorrected for multiple comparisons.