General Linear Model

General Linear Model

• The design matrix, G, can contain both covariates, indicator variables and confound variables reflecting the experimental design.
  • Note that same linear model specified by G is applied to all image voxels, but each voxel can have different b parameters (i.e., regression coefficients).
• Each column in G is associated with an unknown parameter in b.
• Some of the b parameters will be of interest, while others will be of no interest.