afni.nimh.nih.gov

Hello.

When we look at the timecourses of the output from 3dDeconvolve, we end up with many voxels that have beautiful timecourses, but are going in the wrong direction -- their peaks are negative rather than positive. Are these negative-going voxels very common or are we doing something wrong to be getting so many? Is there a way to specify only positive-going voxels in the analysis?

Thanks,
Krista