Generally for tissue volumes(GM, WM and DBM) we use 8-12 FWHM for smoothing. But in some softwares recommend 15 mm for cortical thickness and 20-25 mm for all other indexes (gyrification, cortical complexity and sulcus depth).
Hi Wasana,
Generally, we aim to reduce noise by smoothing and consequenctly increase power. However, aspects of different neural tissue types (e.g. GM, WM) may get combined if brain volumes are smoothed in volumetric space. Since different tissues may differ greatly in function/responsiveness with respect to the measurement methodology, we might not want to mix their properties too much.
But if you are doing an analysis in surface space, one can target one specific type of brain tissue. Under such conditions, smoothing may not blur different tissue properties (although different neural functions may get blurred).
Overall, surface based analyses allow for larger smoothing kernels because of their ability to target specific tissue types.
Cheers,
Wouter
- Badges
- Science topic