Hi,
You just define the kernel but you do not apply it. You have to specify
another argument such as -fmean to do the smoothing. -s is a shortcut
and the results of both ways should be identical.
hope this helps,
wolf
On 3/18/15 12:19 PM, Florence Bouhali wrote:
> Dear FSL experts,
>
> Since my update to FSL 5, my cherished fslmaths command with "-kernel gauss" option decided to go on strike and simply output a copy of the input file, whatever the sigma used.
> This problem has been reproduced with different types of kernels, and on a lab mate’s computer (which came as a surprise since I could not find any mention of such issue on the list).
>
> I have however noticed that the "-s" flag (still in fslmaths) was working.
>
> I thus have two questions:
> 1) Is there an easy fix for this problem or something I may have missed?
> 2) Could someone please confirm that the two commands "fslmaths input -kernel gauss 1 output" and "fslmaths input -s 1 output " should actually perform the exact same computation (or explain the difference if any)?
>
> Thanks,
> Cheers,
>
> Florence
|