Hello
I have only recently started working on FSL and have a basic question on the
statistics. I am only at the beginning of my research venture and have yet
to master the stats involved. For the time being I need to work out a power
of analysis for my groups for my research proposal.
Is there a way I can use the software to compute a sample size needed/ power
of analysis of my planned groups for a fMRI study? I have used FLAME but I
am not sure what stats will be of relevance for a power of analysis for my
initial proposal.
What do people generally do?
Do they use the amount of voxels that activated in a region of interest
above the Z threshold?
What tools can I use to define a region of interest? I am using a Stroop
task and would like to use anatomically pre-defined regions of interest
(Anterior Cingulate etc etc.) first. Thereafter I would like to experiment
by subtracting the Color Naming and Word naming of one sample from the
Interference task from an independent group and use this as a region of
interest.
Would it work say to count the number of voxels in an anatomical area then
compare it to the number of voxels that activated above my threshold in a
post FLAME analysis? Is there an easy way in FSL?
I have yet to attend the FSL workshop as you can undoubtedly see and I am
trying to self train in the mean while. I have run a mean analysis as well
as a T-test successfully, but I am still a bit unclear on these rather vital
points!
Thank you very much!
Stefan du Plesis
CUBIC, Cape Town, South Africa
|