Hello all,
I am attempting a covariate analysis of fMRI data at the second level
(RFX) with three contrast images per subject (10 Ss) and a behavioral
measure (one per contrast image) as the covariate. In order to
specify both subjects and covariate in the design, I am running the
analysis from the PET/SPECT menu as a Multi-subjects: covariates
only analysis. This approach has been suggested for similar
analyses in the list (e.g.,
http://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=ind03&L=spm&P=R440130&I=-1
). Is this a valid analysis in general, and if so, can it be
considered a second level/random effects analysis?
If this is not valid, what is the proper way to ask for fMRI covariates
of per-session or per-condition response to a stimulus? For
example: if chronic pain patients are administered a painful stimulus 30
times on each of 3 different days, we would expect the brain response to
the stimulus to covary with the per-day pain ratings. We want to
remove subject-specific variance and preserve voxels that vary with the
pain rating between the different days in most subjects. We made
contrast images for each subject for each day. Is there some other
way to ask this question?
Thanks for your help,
Brandall Y. Suyenobu, Ph.D.
Staff Research Associate
Brain Imaging Core
CNS: Center for Neurovisceral Sciences & Women's Health
CURE: Digestive Diseases Research Center
UCLA Division of Digestive Diseases
Greater Los Angeles Veterans' Administration Healthcare System
Bldg. 115, rm 223
11301 Wilshire Boulevard
Los Angeles, California 90073
Tel: (310) 478-3711 ext. 40580
Fax: (310) 794-2864
http://ibs.med.ucla.edu
http://mindbody.med.ucla.edu