mat files of slice timing and realignment


Dear DPARSF forum, dear Chao-Gan,

i struggle with the content of several mat files created during preprocessing.

I ran slicetiming and realignment on a 4D EPI dataset and wonder about the .mat files created during that process:
1) both, a*.mat and ra*.mat are located in the FunImgA folder, together with the a*.nii. the FunImgAR only contains the ra*.nii file. is that correct, or should the ra*.mat be also in the FunImgAR folder?
2) is the information stored in the *.mat files also included in the .nii header (rescliced)? i.e. when using the .nii files in spm, do i have to copy the .mat files to the same folder or are the a*.nii and ra*.nii files rescliced and therefore complete?
concerning the values in the *.mat files, i also wonder if my analysis ran correctly (i attached two screenshots)
3) what does the info in the a*.ma reflect? i thought slicetiming only affects slices within each volume and no change is applied across volumes.
4) in the ra*.mat, the first volume is 0, as this is the reference, but all volumes afterwards have the same value? but that means only volume 2 is corrected and all other volumes get the same translational info?
Thanks for any comments!
Best, Val
Image icon a*.mat37.5 KB
Image icon ra*.mat34.25 KB

Hi Val,

1. both, a*.mat and ra*.mat were generated in the step of realign (estimate & reslice).

    a*.mat is generated by "estimate", i.e., the rotation matrices are stored in a*.mat, you can skip the step of reslice.
    ra*.mat is generated by "reslice". Since all the volumes were resliced to the same position of the first volume, they all have the same affine matrix -- the same as the first volume.
2.  For the resliced filles, the information stored in the *.mat files are also included in the .nii header -- the affine matrix is all the same.
3. As in point one, a*.ma is not refecting slice timing, but realign-estimate.
4. I guess is because that the first one is already stored in the 4D .nii, thus the first in .mat is 0. However, you are safe to use the .nii file only, as the resliced .nii file should use the affine matrix in the header only.