Dear Guillaume
That was a great tip. I should have figure this out on my own, but
thanks a lot for your help.
Best
Torben
Sendt fra min iPhone
Den 26/08/2010 kl. 14.14 skrev Guillaume Flandin <[log in to unmask]
>:
> Dear Torben,
>
> not from the interface but you can make sure that the sum, within
> the mask, of the coefficients of the smoothing filter is always 1,
> for the whole image, by dividing your smoothed image by a smoothed
> version of the mask.
> This is done for example in spm_est_smoothness.m lines 219-229.
>
> Best regards,
> Guillaume.
>
>
> On 23/08/10 15:29, Torben Ellegaard Lund wrote:
>> Dear list
>>
>> I have a question about smoothing of images where the background
>> contains Nan values. I can see that it is possible to maintain the
>> Nan voxels by choosing yes under implicit masking in the smoothing
>> job, but is there a way that I can avoid signal loss in edge voxels
>> due to averaging with the Nan values (It seems that in this context
>> they are treated as zeros)?
>>
>> Best Torben
>
>
>
> --
> Guillaume Flandin, PhD
> Wellcome Trust Centre for Neuroimaging
> University College London
> 12 Queen Square
> London WC1N 3BG
|