Fig. 2From: Identification of multimodal brain imaging association via a parameter decomposition based sparse multi-view canonical correlation analysis methodCanonical weights on real data. The top row belongs to SMCCA, and the remaining two rows correspond to the shared and specific results of our method. Within each panel, there are three rows corresponding to three types of imaging QTs, i.e. AV45, FDG and VBMBack to article page