Dear all,
I am trying to unconfound my Nifti data before running MELODIC. This is the way I try to do it:
> fsl_glm -i (nifti file) -d (.txt file with the confounds) -o nifti_betas.nii --out_res=nifti_confoundCorrected.nii
The nifti input file is the file from the scanner, preprocessed (realignmnet, smoothing, coregistration and slice timing) by SPM. The confounds.txt file is basically an M*N matrix with M the number of time point and N the number of confounds. This matrix also includes a constant and a linear and a quadratic trend. When I try to run the command, the new files (nifti_betas.nii and nifti_confoundCorrected.nii) are created but the files are almost empty (zero and four bites repectively). No errors occur. So something goes wrong in the estimation of the betas. Can somebody tell me what is wrong? I think I should pre-process the columns of the confound matrix someway... but I don't quite know how.
Thank you very much!
Regards,
Debby
|