Hello All,

I have some voxels of interest that constitute a mask in Talairach space, which I want to warp back to the original (+orig) coordinates so I can use the mask with either 3dmaskave or 3dmaskdump to extract the time series. I can do this fairly easily within AFNI by setting my 3d+t dataset as my underlay and my mask as the overlay, which I view iin Talairach space, then switching (warping-on-demand) to original space, opening the graph window, and saving the time series ("Opt" --> "Write Center"). This works well and the voxel coregistration (for lack of a better term) between +orig and +tlrc coordinates appears to be reasonable. However, this process is very time intensive due to the sheer number of voxels I have for each subject and the number of subjects I need to do. It would be really nice if I could script this out...

I've tried to use 3dfractionize (template = 3d+t+orig dataset, my underlay; input = mask+tlrc; clip = .20; vote) to simply warp the mask to the individual +orig coordinates, but this doesn't work despite much tinkering with the options: I either get "0 nonzero mask voxels" written to the dataset, or, in the cases where the fractionization process seems to work, I get a "mask and dataset are not the same dimensions" error. I figured this is because the mask is a FIM and the dataset is a 3d+t. So is there any good way to do this, considering I can't use the other AFNI programs that require either a text list of coordinates or the -cmask option in 3dmaskdump?

Thanks!!