tigrlab / compare_task_tools Goto Github PK

Event regressors:
EA video osnets and durations
Circles videos onsets and duration
EA button presses (Dur = 0)
Circles button (Dur=0)
parametric modulators or EA (edited)
1:50
Regressors: mf.CSF mf.WhiteMatter fd mf.X mf.Y mf.X mf.RotX mf.RotY mf.RotZ and then diff for: mf.X mf.Y mf.X mf.RotX mf.RotY mf.RotZ

The GLM t-map contrasts were not consistent in Nistats in comparing to AFNI and SPM, which based on 10 random subjects that we have chose.

What we know right now:

1. By averaging parametric modulation across 6 videos (similar to SPM and AFNI way) and adding these values into the design matrix for GLM analysis. The output t-map contrasts were still not consistent.
2. The columns in the design matrix were highly correlated across SPM, AFNI, and SPM
3. I have rerun the GLM analysis with N-back task, and the results were consistent between SPM and Nistats
4. We decided to use AFNI to run GLM analysis for SPIN EA task

Future thing to do:

1. Post a precise and sophisticated question on neurostars to address how Nistats incorporate the parametric modulation in the GLM model.
2. Perhaps we should share our results so other people who ran into the same problem can read what we have done.