You do not need to re-run the fit. Just run selxavg3-sess with the -contrasts-only option. The new selxavg automatically performs temporal whitening which can make your first-level sig maps appear to be "worse". You can try turning off the whitening with -no-whiten.

doug

Sebastian Moeller wrote:

Hi List, hi Doug,

I just started to use freesurfer 4.5.0 for functional data (before I only used 3.0.5). Now I see that the whole analysis workflow has changed from mkanalysis-sess.new -> selxavg-sess -> mkcontrast-sess -> stxgrinder-sess to mkanalyss-sess -> mkcontrast-sess -> selxavg3-sess, now what happens when I want to add/change contrasts after running selxavg3-sess? Do I still run stxgrinder-sess or rerun selxavg3-sess (can I avoid recalculating the full GLM)? And more interesting, I seem to use mkanalysis-sess.new incorrectly as I get much worse results from the selxavh3 analysis than from the old one. Is there anything special to take into consideration of non human primate analysis (I already fix up the gamma function and autocorrelation timewindows to reflect the use of mion in monkeys). If anyone has a good idea what to try I am all ears...

best Sebastian _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer