Fig. 3: QCAlign verification of regional atlas-registration at the selected intermediate hierarchy level.
From: Detecting the effect of genetic diversity on brain composition in an Alzheimer’s disease mouse model
a Mean accuracy scores per intermediate hierarchy region after QuickNII registration alone (navy triangles) or after QuickNII and VisuAlign registration (green circles). Two raters scored the same 5 randomly selected brains after QuickNII registration alone, max n = 10 assessments per region (Raters: n = 2 per brain). Up to 10 raters scored the same 5 randomly selected brains after QuickNII and VisuAlign registration, max n = 36 assessments per region (Raters: n = 6–10 per brain, see Supplementary Data 2 for the exact number of assessments measured per region). Dark shapes represent the mean score across raters per region for 5 brains +/− SEM, with the opaque shapes representing the individual assessments contributing to each mean calculation. b The impact of VisuAlign refinement on regional stain load (%-stain-positive coverage/per region area) was measured by calculating the difference in load following Nutil quantification after each method (regional (QuickNII + VisuAlign output (%)) – regional (QuickNII output (%)) = regional load difference (%)). Boxplots of individual regional differences in load values are represented for all 5XFAD animals at 6 m and 14 m (6 m: n = 17 mice, 14 m: n = 20 mice). Dots represent mean regional load difference +/− SEM for all 5XFAD animals at 6 m and 14 m (6 m: n = 17 mice, 14 m: n = 20 mice). DG dentate gyrus, Nuc nucleus, Thal thalamus.
