Hello, I am currently working with a functional MRI task involving multiple sessions and multiple runs. All files were preprocessed in Conn toolbox and regression and movement outliers were used. I am currently using the batch system to process these files. This works perfectly for all subjects that have the full 6 runs available. However, I am getting an error in contrasts for any subjects missing any amount of runs.
1) The initial design matrix for these subjects have the gray -> beta not uniquely specified. How would this occur? I assume it is some result of the preprocessing in conn, but am not sure how I could correct it. Would running the preprocess completely separately for these subjects help?
2) Once the design matrix is rank-deficient, I am not able to run the contrasts as before. How would I adjust these contrasts, using the batch system, to be able to run appropriately?
All help is much appreciated, thanks!
|