How Many Subjects Should Be Included in a Well-Powered Cross-Sectional Cortical Thickness Analysis?

Heath Richard Pardoe¹, David F. Abbott¹, Graeme D. Jackson^1,2

A critical aspect of study design is the inclusion of appropriate numbers of subjects. Standard methods for power analysis were applied on a vertex-wise basis to estimate how many subjects should be included in a cross-sectional MRI-based cortical thickness analysis to ensure the study is well powered. Based on results from a standard clinical MRI scanner, we found considerable heterogeneity in the minimum number of subjects required across the cortical sheet. Cortical thickness analyses of the temporal lobe require more subjects for a well-powered study than the frontal, parietal and occipital lobes.