Multiple Subject Barycentric Discriminant Analysis (MUSUBADA): How to Assign Scans to Categories without Using Spatial Normalization

Figure 2

Design and data matrix for the MUSUBADA example. A row represents one image belonging to one of seven categories of images (female faces, male faces, monkey faces, dog faces, houses, chairs, and shoes). A column represents one voxel belonging to one of the ten participants. At the intersection of a row and a column we find the value of the activation of one voxel of one participant (i.e., column) when this participant was watching a given image (i.e., row) Note that the participants’ ROI was drawn anatomically for each participant, so the number of voxels differs between participants (i.e., the scans are not morphed into a standardized space).