Accordingly, this study focused on change in regional brain volumes over an average of 7 months in adolescent boys and girls and identification of image quantification approaches that would be sensitive yet robust enough to detect changes reliably over this short interval. To calculate volume change using atlas-based image segmentation with the SRI24 atlas (Rohlfing et al., 2010), two principal image registration approaches were tested. The “independent” method required two inter-subject registrations to the atlas: both baseline (MRI 1) to the atlas and follow-up (MRI 2) to the atlas. The “sequential” method required one inter-subject registration, which was MRI 1 to the atlas, and one intra-subject registration, which was MRI 2 to MRI 1. We tested the hypothesis that the sequential approach for quantifying regional volume change would yield lower coefficients of variation (CV) of change scores across individuals than the independent method, rendering differences more reflective of true developmental effects rather than misregistration. Considering known growth patterns (Lenroot et al., 2007; Sowell et al., 2004a), we predicted that 1) cortical tissue volumes would decline, 2) ventricular and cortical sulcal
