Figure 1.

Network identification using Multivariate Distance Matrix Regression (MDMR). Cigarette consumption and resting-state functional MRI (rsfMRI) data were collected for each participant (A). For each voxel in the brain, the voxel was used as a seed region to create a connectivity map for each participant (B). These maps were compared with each other to create a subject-wise similarity matrix (C). Daily cigarette consumption for each participant was then combined with the connectivity similarity matrix to produce a pseudo-F statistic, which characterizes how individual variation in cigarette consumption explains individual variation in functional connectivity (D). This is repeated for all voxels (E). Each MDMR voxel-wise result was then combined to produce a map of the ability of the connectivity pattern to predict a cigarette consumption score in each voxel (F). A permutation test of the study subjects' labels can be used to test the significance of this pseudo-F statistic.