Dear Darren,
Thank you for your answer. Determining extend threshold (p-value) in
conjunction analyses (in minimum fields) appear to be a quite tough
issue. My basic problem is may be simpler: I have two groups of
subjects (actually they are the same subjects, but this is two
distinct fmri sessions) and I wanted to know which regions are
involved in the same contrast in both groups. A conjunction analysis
seems appropriate (the two contrasts are orthogonal). My problem is
that as a result I got clusters including e.g. 1, 2, 10 or 50
voxels. Which extend threshold to use to select regions with
significant joint activations? (in the example, 1 and 2 voxels appear
too small, 50 voxels significant, and 10 voxels in between). My first
answer would be to use the same threshold as in single contrast (e.g.
15 voxels, i.e. p<.05), because the conjunction has already taken
into account the joint test (at least on height threshold). However,
it seems to be inapproriate in some instance: Consider for example
that each single contrast activates a cluster including 20 voxels and
the resulting conjunction analysis provides a cluster of 12 voxels
within each single-contrast clusters. Using the same theshold (15
voxels) would reject this conjoint activation, but 12 voxels over 20
voxels which co-jointly activated seems to be quite significant. I
was thinking that in this instance, we indeed use the implicit
assumption that we are testing for joint activations within given
clusters and a standard masking analysis might be more appropriate.
But, the same spatial extend threshold problem appears to occur
again, if I am right. What do you think?
Thanks again,
Etienne.
>Dear Etienne:
>
>The preliminary answer is that spatial extent values are not
>provided because they have not been worked out for conjunction
>analyses. It also sounds like what your trying to work out is how to
>do a conjunction on the differences between groups. i.e. how does
>one decide whether activations seen on a conjunction analysis are
>significantly different between groups.
>
>As far as I know this can't be done in spm. Conjunctions basically
>tell one about main effects. For the interactions one has to do
>either fixed effects or random effects-type analyses. This of course
>leads to difficulties sometimes interpreting results because the
>interactions are not being looked at in the same way as the
>conjunction. Perhaps something is possible with masking, but I can't
>see that this would let you reject the null hypothesis. Hopefully
>there will be a more expert addendum.
>
>Darren.
>
>
>>Dear all,
>>
>>I have a question regarding how to determine p-value related to
>>spatial extension (voxel number) in a conjunction analysis using
>>spm99. I noticed that spm99 does not provide any p-value for
>>spatial extension for conjunction analysis, in contrast to simple
>>contrasts. My understanding is that statistical assumptions
>>underlying computations of those p-values are no more valid in
>>conjunctions of contrasts. Is it right? If so, how can we decide
>>significant spatial overlap between several contrasts. Any
>>insights, help or references adressing this question would be
>>greatly appreciated. Thank you.
>>
>>Etienne.
>>
>>
>>
>>Etienne Koechlin, PhD.
>>--------------------------------------------
>>Cognitive Neuroscience Section
>>National Institute of Health
>>MNB/NINDS/Bdg 10, Room 5c206
>>BETHESDA, MD 20892, USA
>>
>>Tel: (301) 435-1583
>>Fax: (301) 480-2909
>>
>>
>>INSERM U483
>>Universite Pierre et Marie Curie,
>>9 quai St bernard,
>>75005 PARIS, FRANCE
>>Tel: (1) 44 27 37 47
>>Fax: (1) 44 27 34 38
>>---------------------------------------------
>
>-------------------------------------------------------------------------
>Darren R. Gitelman, M.D.
>Cognitive Neurology and Alzheimer's Disease Center
>E-mail: [log in to unmask] WWW:
>http://www.brain.northwestern.edu
>Voice: (312) 908-9023 Fax: (312) 908-8789
>Northwestern Univ., 320 E. Superior St., Searle 11-470, Chicago, IL 60611
>-------------------------------------------------------------------------
Etienne Koechlin, PhD.
--------------------------------------------
Cognitive Neuroscience Section
National Institute of Health
MNB/NINDS/Bdg 10, Room 5c206
BETHESDA, MD 20892, USA
Tel: (301) 435-1583
Fax: (301) 480-2909
INSERM U483
Universite Pierre et Marie Curie,
9 quai St bernard,
75005 PARIS, FRANCE
Tel: (1) 44 27 37 47
Fax: (1) 44 27 34 38
---------------------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|