Scene: Figure S3 Effects of sample size on human group average surfaces

Cerebral Cortical Folding, Parcellation, and Connectivity


Figure S3 Effects of sample size on human group average surfaces

The species differences shown in Fig. S2 in the smoothness of group average surfaces might in principle be due in part to the order of magnitude difference in the number of subjects contributing to the macaque (n = 19), chimpanzee (n = 29) and human (n = 210). To address this issue, we generated two additional group average human right hemisphere surfaces, each based on a different group of 19 subjects, as shown on the far right. These 19-subject group averages surfaces are similar in shape to one another and to the 210V group average (third column) in terms of major gyral and sulcal features, but they each have many local surface irregularities that reflect incomplete averaging of features in regions of high folding variability. The group average macaque and chimpanzee surfaces lack such prominent local irregularities and are closer in shape to the individual subjects. This argues that different numbers of subjects need to be averaged in each species to achieve consistency in the group average surfaces, owing to species differences in the degree of individual variability. Put another way, many more human surfaces must be included in order to generate surfaces comparably as smooth as the chimpanzee or macaque average surfaces.

Surface Mesh:32k fs LR, Species:Human, Species:Chimpanzee, Species:Macaque, Atlas:Yerkes 19, Registration:MSMAll, Other Data:gene