Leveling up with SPM

1

My study has 15 people, 5 conditions (videos being watched), 4 runs/sessions per person, collected while they are in the scanner one time. I use the wonderful fMRIprep for preprocessing. I then mask and smooth the epi images, and process them along with the subject’s T1 image (renaming all files to use shorter names) using a modified script for first-level analysis from Andy’s Brain Blog, which yields (for each subject):

ao_02_1.mat
ao_02_2.mat
ao_02_3.mat
ao_02_4.mat
beta_0001.nii
beta_0002.nii
...
beta_0023.nii
beta_0024.nii
mask.nii
ResMS.nii
RPV.nii
SPM.mat

My problem is, there is an SPM.mat file for each subject, and subsequent (second-level) analyses seem to be expecting some sort of grand SPM.mat file that contains all subjects, conditions, etc.

I seem to be missing a step that is perhaps so obvious it isn’t well documented.

Can anyone point me to a hand-holding tutorial on second-level processing?

Thanks,
John

2

If you did a 1st level analysis, you only have one SPM.mat for each subject. Then you need to run a 2nd level analysis which will produce a “grand” SPM.mat file. I can recommend e.g. these resources: https://www.fil.ion.ucl.ac.uk/mfd/Schedule/Schedule.html
Best,
Ondrej

1 Like