Currently only tensor images in atlas space are computed for each stage, but not for the concatenated/final space (when computing a hierarchical atlas). For the QC of the generated atlas, we need to see the individual tensor images in the final atlas space, so these would need to be generated.
Currently only tensor images in atlas space are computed for each stage, but not for the concatenated/final space (when computing a hierarchical atlas). For the QC of the generated atlas, we need to see the individual tensor images in the final atlas space, so these would need to be generated.