afni.nimh.nih.gov
AFNI Message Board
Dear AFNI users-
We are very pleased to announce that the new AFNI Message Board framework is up! Please join us at:
Existing user accounts have been migrated, so returning users can login by requesting a password reset. New users can create accounts, as well, through a standard account creation process. Please note that these setup emails might initially go to spam folders (esp. for NIH users!), so please check those locations in the beginning.
The current Message Board discussion threads have been migrated to the new framework. The current Message Board will remain visible, but read-only, for a little while.
Sincerely, AFNI HQ
We are very pleased to announce that the new AFNI Message Board framework is up! Please join us at:
https://discuss.afni.nimh.nih.gov
Existing user accounts have been migrated, so returning users can login by requesting a password reset. New users can create accounts, as well, through a standard account creation process. Please note that these setup emails might initially go to spam folders (esp. for NIH users!), so please check those locations in the beginning.
The current Message Board discussion threads have been migrated to the new framework. The current Message Board will remain visible, but read-only, for a little while.
Sincerely, AFNI HQ
| December 10, 2004 09:56AM | |
Hi all
this is just to point out that normalizing the beta coefficients for the regressor of interests to the zero-order baseline coefficient (the constant term) from 3dDeconvolve, while apparently more correct from a theoretical standpoint compared to normalizing using the temporal average of each voxel timecourse, may lead to grossly erroneous results. I noticed this recently while fitting a model with two covariates of interests, and a baseline model comprising a 3rd order Legendre polinomial and the 6 motion parameter: whereas the mean image of the EPI run has values that range roughly from 0 to 1100, the zero-order baseline coefficient from deconvolve ranges from about NEGATIVE 500 to positive 1800. I guess this may derive from some degree of collinearity in the baseline model when you include in it the motion parameters, but it's something that people should really be aware of! (at least before submitting a paper......
giuseppe
this is just to point out that normalizing the beta coefficients for the regressor of interests to the zero-order baseline coefficient (the constant term) from 3dDeconvolve, while apparently more correct from a theoretical standpoint compared to normalizing using the temporal average of each voxel timecourse, may lead to grossly erroneous results. I noticed this recently while fitting a model with two covariates of interests, and a baseline model comprising a 3rd order Legendre polinomial and the 6 motion parameter: whereas the mean image of the EPI run has values that range roughly from 0 to 1100, the zero-order baseline coefficient from deconvolve ranges from about NEGATIVE 500 to positive 1800. I guess this may derive from some degree of collinearity in the baseline model when you include in it the motion parameters, but it's something that people should really be aware of! (at least before submitting a paper......
giuseppe
| Subject | Author | Posted |
|---|---|---|
|
Giuseppe Pagnoni | December 10, 2004 09:56AM |
