Among these tests, we demonstrated that different, widely-used programs to measure regional brain volumes produced consistent results on a broad range of cohorts worldwide, based on data from young and old samples and mixtures of both (Stein et al. 2012; see supplement comparing FSL and FreeSurfer segmentations). It was not feasible to require all sites to use one specific program to measure brain regions; as might be predicted, no single algorithm performed best for quantifying brain volumes across all datasets.
