Dear Karl,
I (and perhaps the list) am hoping you could expand on your thoughts
with respect to sources of 'global effects' in fMRI that should be adjusted
with proportional scaling? My understanding of the origins of ANOVA and
proportional scaling was for PET and SPECT imaging where radioactive counts
could be dramatically different between successive scans - and since it was
argued that noise and signal scaled proportionally - these within subject
scaling procedures were employed. However, as PET camera's have become
increasingly sensitive to local signal, they too have run into the
local/global signal confound (as shown in Jesper Andersson's (1997),
Neuroimage, 6(4), 237-44, when proportional scaling/ANCOVA models are used.
I am aware of some MR scanner related artifacts (namely rf) that can lead to
occasionally 'bright' or 'dark' images (i.e., global effects) - but these
are typically uncorrelated with the task making them less likely to be
susceptible to the problems outlined by Aguirre et al in (1998) with
proportional scaling. Also, I tend to see these artifacts less and less as
MR equipment matures for fMRI protocols.
I am familiar with your web site's description of proportional
scaling (http://www.fil.ion.ucl.ac.uk/spm/doc/intro/#_E__Spatial_smoothing).
In particular, the web site states that 'However, the issue of
normalization-induced deactivations is better circumnavigated with
experimental designs that use well-controlled conditions, which elicit
differential responses in restricted brain systems'
However, it is not always possible to design studies that have
well-controlled conditions, such as the case of Go/No go paradigms (where
the substantial motor activity from one condition (Go) is more highly
correlated with the global signal than is the inhibitory activity associated
with the 'No go' response. Thus, proportional scaling runs amuck in these
situations.
I guess my longwinded question comes down to this:
If we assume we can control scanner related 'global' signal confounds, is
there any reason to still use proportional scaling? (Even if the conditions
in the paradigm are perfectly balanced)?
Best regards,
Kent
-----Original Message-----
From: Karl Friston [mailto:[log in to unmask]]
Sent: Monday, March 21, 2005 6:54 AM
Subject: Re: Global Scaling
Dear Bo,
>now, another basic question. how to judge whether I should chose
>"none" or "scale" in fMRI->data. how to know whether the global brain
signal
>was correlation with my task? I get the rd_taskcorr.m file from web which
>could computer the correlation P value, rewrote it for SPM2 and test my
data.
>I found in same one study, some subjects could be scaled, and others could
>not. what's your opinion about this question?
Personally, I would usually use scaling. The reason is simple; If you
are interested in regionally specific effects, then, by definition global
effects are confounding (irrespective of whether you can detect them
or not). I appreciate some people find it difficult to interpret models
with global normalization in absolute physiological terms but these
difficulties have to be seen in the context of partitioning responses into
regional and global components It is perfectly OK to do a T-test (i.e.
a correlation) on the global activity and an SPM{T} of regional effects.
In designs with low-level baselines and widespread activation one can
sometimes
get a significant global effect and no regional effects. This may seem
paradoxical from a physiological point of view but is quite proper in
relation
to the differences between global and regional effects.
There are many qualifications here. For example, two regions may express
a task-related response that is highly correlated with global activity.
However,
one of the regions expresses it with twice the amplitude. Is this a regional
effect or a global effect?
With very best wishes,
Karl
