> We want run the segmentation from the command line and we intending to do
> multispectral segmentation.
>
> 1. How do I specify the input images to the algorithm(spm_segment.m)? I've
> three images(T1,T2 and PD) for each subject.
P=spm_get(3,'*.img','Select images for this subject')
and then pass P to the program.
> 2. Do I need to use the spm_vol.m function before I use the spm_segment.m
> function?
You should be able to just pass the filenames.
> 3. When I see using breakpoints how the inputs from the GUI are taken,the
> input to the segmentation algorithm seems to be a structure with several
> fields. Do I need to specify all the fields?
It should run without having all the fields passed. If you want to pass them,
then all you need to do to get the flags is:
spm_defaults
global defaults
flags = defaults.segment;
>
> I tried most of the ways possible but I just can't understand how to do it.
> Can someone please help!The reason I want to do it command line is I'm
> writing a script for batch processing of several subjects with our own
> templates and own priors.
Best regards,
-John
|