Print

Print
Hi,

Yes - you can do this with voxelwise EVs in randomise.
Just use the --vxl and --vxf options.
The regressors in this case, naturally enough, have to be stored as 4D nifti files and the input to --vxf specifies where to find these files and the input to --vxl specifies which columns of the design matrix they correspond to.

All the best,
Mark



On 14 Nov 2018, at 21:32, Eckhard Schlemm <[log in to unmask]> wrote:

Hi,
I understand how to use randomise, cluster and fsl_glm to perform a mass-univariate regression of a vertex-varying response variable against a global predictor, followed by, e.g., cluster-extend based thresholding to identify a FEWR-significant cluster. I believe the canonical example of this kind of analysis would be correlation of cortical thickness with age.

Is it possible in FSL to have a predictor that varies across vertices, and do a correlation analysis of, e.g. cortical thickness against cortical curvature? I presume that this would necessitate specifying a vertex-dependent design matrix in the GLM.

Thanks,
Eckhard

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1


To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1