Figure 2
Cross-modal affine image registration using vessel branchings as fiducial points and VBM analysis workflow. (a) Vessel branchings are visible in MRI as well as 2Pii without any additional contrast agent. Upper left panel shows view on exposed brain surface after window surgery, regions 1–3 indicate the positions of vessel branches in MRI and 2Pii. (b) Six partially overlapping 2Pii-stack-masks of one animal registered on the corresponding MR image. Average intensity projections; up: horizontal view, bottom: sagittal view. Scale bar = 2 mm. (c) Longitudinal registration and processing of MRI 3D images for VBM. Longitudinal nonlinear registration was performed for each subject between timepoints (t1 = baseline, t2 = 1 week or 12 weeks), which yielded resulting average and Jacobian Rate (JR) images. The average images were pre-segmented in three tissue probability maps [gray matter (GM), white matter (WM), cerebrospinal fluid (CSF)] and multiplied by the JR-images. From these, a Dartel prior knowledge template was created and the resulting flow fields were used to normalize the JR-modulated tissue maps into Paxinos space. They were finally smoothed and statistically analysed in SPM second-level models.
