 |
 |
Dear Simona, > Dear SPMers. I am running a VBM analysis following the pre-processing > procedures reported on the SPM website > (http://www.fil.ion.ucl.ac.uk/~john/misc/VBMclass10.pdf). I would like to > extract the GM volume from a region of interest based on MNI coordinates > using the pre-processed images. These values will be used as covariates in > my statistical model. With previous SPM versions, I used to extract this > value from modulated, normalized scans. However, 'DARTEL Normalize to MNI' > step only produces smoothed images. Which images should I use to extract > these values? Is there a procedure to extract the values from the segmented > original scans? I think you are still fine to extract the values from the DARTEL-normalized images in MNI space. First, you can specify 0mm smoothing - the results will look a bit odd because of aliasing (see many previous posts about this), but I don't think this will harm your results. You are averaging over an ROI, anyway, so what is of interest for you is that the *average* values for this region are correct. Second, if the aliasing is a concern, you could also smooth by a small amount (e.g., 1 mm), which is unlikely to have a huge impact on your results (depending on the size of your ROI). Finally, though, you may want to consider whether you actually want to use smoothed data, anyway (i.e. with a FWHM of 6-8 mm). Although high-dimensional registration algorithms such as DARTEL are very good, they remain imperfect. Further, even if the spatial normalization of brains were perfect, the correspondence between cytoarchitecture and the macroanatomical features apparent on T1 scans is imperfect. Thus, if you are really trying to measure some type of common structural region which varies across participants, spatial smoothing is one way to potentially increase the accuracy of your measurement. All of this, of course, depends on the particulars of your data and the assumptions you are going in with, but hopefully one of those options will work for you. Hope this helps! Best regards, Jonathan -- Dr. Jonathan Peelle Center for Cognitive Neuroscience and Department of Neurology University of Pennsylvania 3 West Gates 3400 Spruce Street Philadelphia, PA 19104 USA http://jonathanpeelle.net/