On 30 August 2010 09:46, B. Martines <[log in to unmask]> wrote:

Hi,
I have fmri-data with event-related design. I want to analyze for functional connectivity using PPI. Can i use Beta-Values instead of time course? And how i get Beta-Value using a ROI?

I think it's feasible, although you'd want a lot of events, and the interpretation will be somewhat different. You need to generate a separate regressor for each event, and then use fslmerge to merge the appropriate PE files from the stats directory into a 4D file.

Eugene

Thank you very much for your help

Martines