Figure 3.

Ground‐truth (HR T2W) similarities for static and reference‐guided SRR outcomes for the quasi‐static control brain experiment whereby each of the seven subjects is assigned a different marker. The more input stacks are used the higher the similarity scores. Moreover, motion correction markedly improves the ground‐truth similarities which was performed by rigidly registering each individual slice to the HR T2W volume using NCC as the similarity measure. A visual comparison for one subject is provided in Figure 4. Stars indicate statistical differences between the groups using a pairwise Wilcoxon signed‐rank test (p < 0.05)