Figure 4.

Comparison of the dyslexic subjects' brain classifications for data smoothed with a 4, 8 or 12 mm FWHM isotropic Gaussian kernel (from left to right). For the three sizes, significant clusters of differences were obtained in the right cerebellar declive (corrected p values ~0; cluster size under H0 for voxels with 100% of difference = 0.36 +/- .05 at 4 mm smoothing, 0.531 +/- 0.07 at 8 mm smoothing, 1.33 +/- .09 at 12 mm smoothing). By contrast, only one cluster was observed in the right lentiform nucleus with a 8 mm smoothing kernel vs. 1 voxel only with a 4 mm or 12 mm smoothing kernel. As illustrated, the coordinates of the centres of mass (or single voxel) differed with the smoothing kernel. However, by lowering the threshold of the PMD from 100% to 95%, there is a clear overlap between the three PMD (right hand side), confirming the robustness of the results.

Pernet et al. BMC Neuroscience 2009 10:67   doi:10.1186/1471-2202-10-67
Download authors' original image