Hi,
this only means that the map is significantly, but negatively
correlated to the associated timecourse. Inverse the timecourse, and
the voxels were painted red. This has been discussed some time ago,
but right now I can't find the thread. Browse the FSL list if you need
to actually inverse timecourses etc. For spatial purposes, this should
not matter AFAIK.
Cheers,
Cornelius
On Thu, Jul 7, 2011 at 6:27 PM, Stuart Washington <[log in to unmask]> wrote:
> To whom it may concern-
>
> I am a regular user of MELODIC but this is my first question to the group. I have performed ICA on a resting state data set and it yielded 18 maps. One coherent resting state network component (which we think is auditory) is visible ONLY when a -3 (min) and -8 (max) threshold is used. All other network/motion/arterial components (including visual, DMN, motor, etc) are visible when thresholded from 0 (min) to 8 (max). I've seen this same supposed-auditory network in the positive range in other data sets.
>
> Given that this is functional connectivity (i.e., how correlated voxels are with one another) and not activation (i.e., how active voxels are relative to rest), is it even significant to my results that this map is negative? I am currently writing a paper on DMN and need to compare it with other networks so it is critical that I know what is going on here.
>
> --Stuart
>
--
Dr. med. Cornelius J. Werner
Department of Neurology
RWTH Aachen University
Pauwelsstr. 30
52074 Aachen
Germany
|