Hello!
I am using the preprocessed rsfmri data of the human connectome project. We used '.pconn.nii' files containing cross-correlation matrices on PALM. Earlier palm didn't have the functionality of'pconnscalar.nii' where we can enter 379*379*N , where N being the no. of subjects. I have certain doubts regarding the approach we have adopted and whether it is correct or not. Could you answer the following?
1. Earlier we read the cifti file containing cross-correlation matrices and created a .csv file containing correlation b/w parcel A/1 with the rest of the 379 parcels for N subjects. The dimension of the file is in the format N*379 where N is the no. of subjects. We created 379 .csv files for 379 parcels and used it to run a GLM model using the below line of code:
palm -i parcel1.csv -d design.csv -t contrast.csv -eb EB.csv -n 5000 -accel tail
Here contrast contains both positive & negative values.
i). Is approach 1 correct if we are aiming for an exploratory analysis? If it is, is there a way to automate it on matlab?
ii) Does entering the two contrasts in one model make it a two-tailed test? If not, how can I do it?
iii) Could you share how FWE correction is being applied here?
2. With the new functionality, when I entered 379*379*N pconnscalar.nii file, no region survived FWE correction.
i) Could you explain how the correction is being applied here?
ii) How the correction being applied in 2 is different from 1?
iii) How to apply Dunn-Sidak correction here?
iv) If I enter a 10*379*N matrix, where 10 represents the no. of a priori parcels whose connectivity I want to check with rest of the 379 parcels, will this approach be correct?
3. To compute interaction, do I just need to multiple the 2 scores for which I want to check interaction?
Thank you in advance!
Regards,
Shruti
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=FSL&A=1
This message was issued to members of www.jiscmail.ac.uk/FSL, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|