Hello,
As this is just for visualisation there is no right answer here; fslmaths takes the gaussian kernel size in sigma, 2.12322 corresponds to a FWHM of ~5mm, while 0.125 will obviously give a much smaller kernel. Without knowing the size of your data I cannot recommend a specific kernel size, but if 0.125 empirically provides the optimum visualisation experience then I would stick with that value.
Hope this Helps,
Kind Regards
Matthew
--------------------------------
Dr Matthew Webster
FMRIB Centre
John Radcliffe Hospital
University of Oxford
> On 28 Dec 2017, at 04:35, Josue Avecillas <[log in to unmask]> wrote:
>
> Dear Experts,
>
> I am using Gaussian kernel function of the fslmaths to smooth some probabilistic maps for visualization. In previous post I've seen a recommendation of using the following command "-kernel gauss 2.1233226 -fmean".
>
> I am not familiar with the concept of spatial smoothing but changing the numbers from 2.1233226 to 0.125 gives me nicer maps in terms of visualization. Is it correct to use "-kernel gauss 0.125 -fmean"?
>
> Thank you so much
|