----- Original Message -----
From: Douglas Burman <[log in to unmask]>
Date: Tuesday, March 15, 2005 4:50 pm
Subject: Re: [SPM] Any Papers on Presenting fMRI Results?

> Daniel and Andrew,
>
> I mostly agree with your points also, but just to make a small
> clarification of my earlier statement...
>
> >Doug Burman wrote a very helpful message that concluded:
> >
> > > So neuroimaging does not localize function.  What it does do is
> > > indicate that the activity of some areas are more involved in some
> > > types of activities or functions than others, which allows
> > > inferences to be made.
>
> To clarify... neuroimaging allows us to determine that "...the
> activity of
> an area is more involved in some types of activities or functions than
> other types of activities or functions."
>
> If  1) motor cortex is experimentally shown to more involved in motor
> planning than visual activity and  2) visual cortex is shown to be
> moreinvolve in visual processing than motor planning (both
> demonstrable by
> appropriate experimental designs), then it is a reasonable
> inference that
> motor cortex is preferentially involved in motor planning.  It is not
> correct, however, to infer that visual cortex has no role in motor
> planning(even though we may reasonably infer that visual cortex is
> more directly
> involved in visual processing), nor can we assume that motor cortex
> has no
> other role.

This hits the nail on the head I think. What you describe is almost a
weak sort of double dissociation, and is even more compelling when both
findings are part of the same study. Because it's possible to show more
visual cortex activation in one condition than in another, and the
inverse for motor cortex, one can reasonably conclude that sufficient
statistical power exists in both areas to detect an effect of a given
strength. And because we're directly comparing two conditions that
activate two regions preferentially, we can conclude something about
localisation, though conclusions about exclusivity still cannot be made,
as you say.

> Broca's area is another great example.  Broca's is preferentially
> activatedby language compared with many other types of tasks, but
> under the right
> circumstances it can also be activated by music or by learning
> paradigms.  Broca's involvement in language may be reasonably
> inferred, but it cannot be claimed that Broca's function is
> exclusively linguistic.
>
> Doug Burman

One aspect of the discussion on thresholded versus unthresholded maps
that has been more or less lost in the to-and-fro is whether images
should use a colour scale to depict a statistical quanitity (e.g. Z or
T), or whether a unit more related to magnitude of signal change (e.g. %
signal change) should be reported (this applies equally to graphs of
extracted cluster activation). I would argue that the latter is a better
technique, because it gives us some indication of whether the failure of
a region to achieve statistical significance has more to do with
variability, or with lack of signal change.

Thoughts?

Tom J.