basis_normall

AFNI Message Board
1

AFNI version info (afni -ver):
Precompiled binary linux_rocky_8: Oct 12 2023 (Version AFNI_23.3.01 'Septimius Severus')

Dear Colleagues,

Hi. I have some good news. Sara Inati, the rest of the team, and I successfully completed our first EEG-FMRI epilepsy nap study. I will soon use AFNI's basis functions to perform an event-related analysis of interictal epileptiform discharges. I have some questions about the extremely helpful basis functions hands-on exercise.

In

https://afni.nimh.nih.gov/pub/dist/edu/data/CD.expanded/afni_handouts/afni06_decon.pdf

on page p. 18, a

-basis_normall 1 \

option is included in the 3dDeconvolve command. However, when I ran the afni_proc.py command in

https://afni.nimh.nih.gov/pub/dist/edu/data/AFNI_data2/s1.afni_proc.command

the option was not included in 3dDeconvolve in my proc.ED.8.glt script. If I want the beta coefficients to represent the percent signal change, should I add

-regress_basis_normall 1 \

to my afni_proc.py command, and what does the following in the help for afni_proc.py mean? "For the example basis functions, -basis_normall is not recommended."

Sincerely,

Dante

2

Hi Dante,

Congrats on completing the nap study.

Interesting, I never noticed -basis_normall in that particular command. It has no use or effect being there. That option will have 3dDeconvolve scale all subsequent regressors to unit height. But TENT regressors are automatically scaled to be that.

It was likely copied from a more appropriate 3dDeconvolve command.
I will remove it.

Thanks,

  • rick
1 Like
3

Rick,

Ah. Thank you for that reply and education.

Sincerely,

Dante