Dear Mariam,
The SPL is an anatomical label, whereas the Anatomy Toolbox (http://www.fz-juelich.de/inm/inm-1/DE/Forschung/_docs/SPMAnatomyToolbox/SPMAnatomyToolbox_node.html ) aims at cytoarchitectonics (although it also provides anatomical labels from other brain atlases, which are not probabilistic though). Thus, if you are interested in anatomy, I'd go with one of the probabilistic anatomical brain atlases like n30r83, LPBA40, Harvard-Oxford. All three come with a maximum probability map (~ voxels are assigned to anatomical labels in an all-or-none fashion), and at least the n30r83 and LPBA40 also have probabilistic labels for each of the different anatomical regions. Intensity range for the probabilistic labels is 0 to 1 for the n30r83 (at least as far as I remember) and 0 to 10000 for the LPBA40 (in contrast to what is stated at http://www.loni.usc.edu/atlases/Atlas_Methods.php?atlas_id=12 ).
When working with probabilistic labels you could weight your data accordingly so that voxels with a low probability are weighted to a lesser extent. This can be achieved with Imcalc i1.*i2 (in case the voxel resolution of the brain atlas differs from that of your data go with a Coreg: Reslice step to reslice the atlas with a voxel resolution matching that of your data, choose nearest neighbour during interpolation). Alternatively, you could also threshold the labels with Imcalc, something like i1>.8 or i1 > 10000*.8, and most straightforward, just go with the labels as included in the MPMs.
Best
Helmut
|