Dear clever people,
Background:
I am conducting a study where 10 participants have been asked to perform a simple tone-paced joystick movement task with their eyes closed. It is an epoch-related the design, where the movements are performed during the six 30 second movement epochs per scan, each subject having two scans during two different conditions (1 and 2). I have created a T-contrast showing the main effect of movement regardless of condition, and at p<0.05 (FDR correction), I have defined a motor network associated with the task. My aim is to now do small volume analysis of each of the regions defined by my motor network, on contrasts examining the difference in scan conditions (T contrasts 1>2, and 2>1).
Unfortunately, in the main effect of movement contrast, I have one large cluster (3213 voxels) which spans a number of motor areas associated with the task (M1, S1, lateral premotor area, superior parietal lobule and intra-parietal sulcal cortex), which SPM has given me peaks in just 3 of these areas. The other regions are also clearly active, but in order to carry out my restricted volume analysis, I need to determine the sub-peaks within the large cluster.
Question:
Is it possible, and is it justifiable, to create a mask of the large cluster and then perform a restricted volume analysis on the main effect of movement contrast to get SPM to give me the sub-peaks, and then use these peaks in the second part of my analysis?
|