Dear Steve,
If you still observe the same value despite making sure that
interpolation is not the issue here as suggested by Pieter and Volkmar,
then it is likely that the issue is a consequence of the automatic
scaling of the data to maximise its range within the data type.
You can change this behaviour by editing spm_imcalc.m at line 149:
https://github.com/spm/spm12/blob/r7771/spm_imcalc.m#L149
so that it uses:
'pinfo', [1 0 0]',...
or, if you don't want to edit an SPM function, just call spm_imcalc by
passing Vo as a struct instead of a filename.
Best regards,
Guillaume.
On 04/03/2020 12:11, Steve Petersen wrote:
> Dear SPM experts,
>
> I tried to multiply a MNI binary lesion mask (1= lesiones, 0 = not
> lesioned) that has the same dimensions as the aal.nii atlas image. By
> multiplying both images (1 x aal labels) I would expect to change the 1
> values of the binary lesion mask for those aal regions where are lesions.
>
> I tried to do this using the following expression in the SPM12
> ImCalc: *i1.*i2. *However the output image not contains the integer
> numbers expected and, for example, a region that should have the 37
> value, contain values with decimals (eg. 37,002). What could be the
> problem?. Is correct the expression used in the ImCalc, I used the
> default options in the ImCalc but should I change some option of the
> Batch editor such as the interpolation method, the data type...
>
>
> Many thanks in advance.
>
>
> Best regards, **
--
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
UCL Queen Square Institute of Neurology
London WC1N 3BG
|