Thank you for providing a wonderful software.
I would like to know how the 6 values in HeadMotion.mat/RealignParameters
were computed for each subject.
I find that the motion derived in dparsf appears to be different that the motion parameters that SPM derives in standard preprocessing. Why would that be? For example, if I view the rp file created by spm in preprocessing there is no motion over 2mm, but if I runt he same data in dparsf, it will suggest I exclude several people at the 2mm cut off. Thanks!
The rp*.txt (head motion parameter file) generated by SPM and by DPARSF are exactly the same, since DPARSF is based on SPM (i.e., call SPM functions to do realignment).
How do you determine there is no motion over 2mm?
DPARSF generated that report by checking the maximum motion in any translation and any time point, if that number is bigger than 2mm. (So as the rotation > 2 degree, if you are looking at the 2mm and 2 degree line).
If you are using DPARSFA new version, you can find the description in a file called HeadMotion.csv.
If you are using DPARSF basic edition, this is the maximum motion for each translation (in mm) or rotation (in degree) (each column) for each subject (each row).