"The most common approach for exploratory ROI analysis is to create small ROIs (usually spheres) at the peaks of activation clusters; in the case of large clusters, it can be useful to create ROIs for additional local maxima in order to explore multiple regions within the cluster... this approach can be very useful for exploring patterns of activity across conditions."
Emphasis is on exploratory analysis, and for exploratory purpose you can do anything. The idea of spheres is that this way, all the ROIs have (approx.) the same size, so you average across the same number of voxels. Depending on point of view this might be advantageous or not. You could also argue that if some anatomical or functional region is indeed larger, then you should also average across this larger volume because you're interested in the behavior of that particular region.
"it can sometimes be equally useful for determining the reasons for lack of activation. As an example, a recent correlational analysis between fMRI and behavioral data in our laboratory failed to uncover any activation in several regions of prior interest, much to our surprise. We performed an ROI analysis on several regions (using small spheres placed in anatomical ROIs) and quickly saw the reason for this lack of activation: Whereas the group as a whole showed a striking correlation, one subject was an extreme outlier who suppressed the correlation."
This is actually bad science in my opinion. >>>After failing to find any sig. correlation<<<, they start to wonder whether there are outliers or not. What about if the correlation had been sig., maybe even due to the outlier? Then you wouldn't check for outliers? And what does it mean that someone's data for certain coordinates, averaged across a certain volume is an outlier? Maybe it's no outlier on voxel-by-voxel basis?
|