Fig. 2.

Weight maps for GPR model predicting STAI-T based on patterns of activation to dynamic emotional face processing using a two-folds cross-validation framework on the whole sample. A: Voxel-based predictive pattern. The colour bar indicates the weight of the voxels for decoding the clinical score. B: Region-based pattern localization map computed from the voxel based predictive pattern displayed in Fig. 2A. The colour bar indicates the percentage of the total normalized weights that each anatomically labelled region explains.