Dear Matthieu,
you cannot directly compare the smoothing sizes between volume and surface because it seems that the effects on the surfaces are somewhat larger compared to that of volumes. According to the matched filter theorem you should filter in the range of the expected effects. However, to be honest I have no idea why the effects on the surface are often larger...
Best,
Christian
On Fri, 26 Jul 2019 15:39:57 +0200, Matthieu VANHOUTTE <[log in to unmask]> wrote:
>Dear CAT12's experts,
>
>In CAT12 manual, it is recommended to smooth cortical thickness maps
>between 12-15 mm FWHM before statistical analysis.
>
>However, in my particular case I need to have an equivalent effective
>smoothness between both PET and MRI imaging. Knowing that PET data are
>smoothed using a 8 mm FWHM, MRI should then be smoothed using a 9 mm
>FWHM according our scanners' characteristics.
>
>Would it be problematic to smooth cortical thickness surface data with
>this kernel size despite the manual recommendations ?
>
>Best,
>
>Matthieu
|