Fig. 1.

Overview of the brain-predicted age analysis pipeline. Illustration of the methods used to generate brain-predicted ages. 3D T1-weighted MRI scans were segmented into gray matter (GM) and white matter (WM) before being normalized to common space using nonlinear image registration. Normalized GM and WM images were concatenated and converted into vectors for each participant. These vectors were then projected into an N × N similarity matrix based on vector dot-products. (A) Once in similarity matrix form, the training participants' data were used as predictors in a Gaussian Processes regression model with age as the outcome variable. (B) Model accuracy was assessed in a 10-fold cross-validation procedure, comparing brain-predicted age with original chronological age labels. (C) Model coefficients learned during training were then applied to the data from DS participants and controls to generate brain-predicted ages. (D) A metric to summarize the variation in brain-predicted age was defined; the brain-predicted age difference. Abbreviation: brain-PAD, brain-predicted age difference.