>Dear SPM Gurus,
>We would like to determine the correlation between increased rCBF in a
>distinct brain area (voxel) and other activated areas in a current
>PET study (e.g like in Toni et al. "Visually instructed and visually
>guided movements", NeuroImage 1998:S980).
>How can this be done?
>best wishes,
>F. Munz
Dear Frank
What you describe is called a "physiophysiological interaction". If instead
you would like to compare correlations between rCBF in two voxels in
different psychological contexts (or conditions) this is called a
"psychophysiological interaction" analysis. For definitions and concepts
see Friston, K.J., Buechel, C., Fink, G.R., Morris, J., Rolls, E. & Dolan,
R.J. 1997. Psychophysiological and Modulatory Interactions in Neuroimaging.
Neuroimage 6, 218-229
The following is a copy of an email on psychophysiological interactions
written for the SPM helpline a few weeks ago. For physiophysiological
interaction, simply use activity in one voxel as the covariate of interest
in Step 5. The results of the SPM represent voxels whose activity
significantly correlates with activity of the specified voxel.
Firstly your original experimental design should be, generally speaking,
factorial.
The following example is a series of steps in the analysis for fMRI
time-series, and can be performed on an individual subject's data. If you
are working with PET scans you generally use multiple subjects to
ensure sufficient degrees of freedom.
Step 1
>From your original SPM factorial analysis, go into Results, choose the
contrast from the main effect of one factor (A1 + A2) - (B1 + B2) in
which there is significant activation (A - B) in the region that you
wish to use as the regressor in the psychophysiological interaction
analysis, and select the voxel of maximum intensity in this area. For
example, if you predict that SI will regress on activity in the
thalamus in context 1 but not context 2 select the maximum intensity
voxel in the thalamus that is active in (A1 + A2) - (B1 + B2) in your
original analysis.
Step 2
Plot activity in this voxel as a function of scan or time. This will
give you a vector (called y in matlab) that corresponds to BOLD/blood
flow in this region (adjusted for confounds) in each scan over the time
course of the experiment. In your matlab window mean correct and save
vector y (i.e. y = spm_detrend(y)). y now 'stands in' for Factor A,B
which should interact with factor 1,2 to explain variance in regions
that evidence an interaction. Factor 1,2 is the psychological factor
and y (A,B) is the physiological factor. A psychophysiological
interaction (PPI) simply means that the ability of y to predict changes
in the areas identified by the PPI depends on the context 1 or 2.
Step 3
Next you need to create a regressor P corresponding to the
Psychological Factor 1,2. It is important to retain the same order of
scans that is assumed by y. So you need to give the value 1 to all
scans corresponding to conds A1 and B1, and -1 to all the other scans
(conds A2 and B2). You should now have a vector (of equal size to your
vector from step 2; length = number of scans in your study), consisting
of 1s and -1s. This vector has a mean = 0 so does not require
mean-correction. This is Factor 1,2 say P = [1 -1 1 -1 ...]';
Step 4
Multiply these 2 vectors together and mean correct the result. PPI =
P(:).*y(:);
Now you have three vectors:
a) activity in a particular voxel over the time course of the experiment y
b) condition order P
c) the interaction between activity in a particular voxel and the
corresponding task PPI
Step 5
Go into SPM, carry out a statistical analysis specifying User Specified
type of analysis. Specify 1 covariate of interest, and enter the name of
the matlab vector that corresponds to PPI above. Enter the two other vectors
(P and y above) as confounds. Enter 2 contrasts: 1 and -1.
This analysis will calculate regressions at every voxel in the brain
for the confounds P and y. Differences between the regression slopes of
BOLD/rCBF on y under levels 1 and 2 of the psychological factor are
tested directly to produce a SPM{t}. The resultant MIP demonstrates the
presence of significant psychophysiological interactions i.e. voxels in
which the contribution of the thalamus changed significantly as a
function of context (1 vs. 2): contrast 1 shows areas for which there
is a significant positive regression slope difference; contrast -1
shows areas for which there is a significant negative regression slope
difference.
Hope this is clear and helps you with your analysis.
Best wishes
Sarah Blakemore & John Morris
*******************************************
Sarah-Jayne Blakemore
Wellcome 4 Year PhD Programme in Neuroscience
Wellcome Department of Cognitive Neurology
Institute of Neurology
University College London
12 Queen Square
London WC1N 3BG
UK
Tel: 0171 833 7472
Fax: 0171 813 1420
http://www.fil.ion.ucl.ac.uk/
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|