Dear SPM experts,
I am using the gPPI toolbox for the first time, and I'm uncertain about the first-level output. The original (non-PPI) model had three regressors/conditions, so my PPI model has seven (in order: the three original, the three PPI, and one physio (and intercept)). In accordance with the PPPI manual I've specified a 'P.Contrast', second condition minus the first, and get as output con_PPI_New.img. Am I correct in that this simply corresponds to subtracting the beta.img for the first condition's PPI regressor (so fourth column in the SPM design) from the second condition's PPI beta.img (fifth column) and this therefore represents the connectivity with the seed region specific for this condition (or at least relative to the first condition)?
Can I now simply continue by averaging over these contrast images from within-subject runs (each subject has three/four runs), and then take these mean images to group analysis? Further, is there any objection to taking these (mean) con images 'outside SPM', and running for instance FSL's randomise on it for group-level?
Thank you very much for any feedback. Kind regards, Caspar
p.s. one more question... Instead of specifying the path to an image for the 'P.VOI' variable, I'd like to use a txt file with values representing the already calculated mean time series from a ROI. Using P.VOI=importdata('timeseries.txt') gives an error ("VOI file is not an .img ..." ); there is something in the manual about PPPI(P, tsdata) but I can't get it to work, either 'too many input arguments' or 'VOI file doesn't exist'. How do I need to run this variant?
|