Hi SPM experts,
I have two separate sets of questions:
1. Questions about head motion plots:
I have two sets of preprocessed data that used the same raw input files, that I have used to generate head motion plots (using "rp*.txt" files). I expected the sets of head motion plots to match given the same raw input, but they did not. Is there an SPM-related reason this would happen? I used SPM 8 for one preprocessing script, and SPM 12b for the other--would that be sufficient to account for the difference in the rp txt files? Or is it more likely that I made a mistake in one of the sets of preprocessing steps if such differences in the plots are present? (Reference slices and other relevant parameters in the slice timing portion of the the script were the same)
Something that might be relevant in considering this is--even though head motions plots look different, they often have matching elements. For example, the corresponding plots often seem like a time-shifted version of each other, or have sections that look like they are mirrored and transposed from one plot to another, or the relationship between the "x" "y" and "z" lines are shifted but some similar patterns remain. For some of the plot pairs, the correspondence is very close, and for others it is barely distinguishable but still present
2. Questions about masking out CSF and white matter on the second level:
I have a set of data that has gone through preprocessing and first level analyses, and would like to mask out white matter, CSF, and ventricle activation in conducting my second level analyses. I have read that masking is possible on the second level, but do not know how I would go about doing so or acquiring an image to use for masking.
One method for creating mask images that I have read about involves using images created during the "segment" step of preprocessing. However, my preprocessing did not involve a separate segmentation step--I used the new "Normalize" step in SPM12b which I was advised was superior to using the old normalization, which would have involved a separate segmentation. My questions are as follows:
a. Ideally, I would like to find a way to use a mask excluding white matter/ventricle activation/CSF not on every subject, but on the actual template that the second level activation differences are being projected on to. Is this possible? If so, how would I apply such a mask?
b. If that is not possible, how can I create masks for each subject to use at the first level step if my preprocessing has already been run without segmentation? And, once I have these images, where would they be applied?
Thanks in advance for your help!
--Victoria
|