This toolbox was developed by Alfonso Nieto-Castañon and Ev Fedorenko to facilitate the process of conducting subject-specific analyses in SPM.
The toolbox should be compatible with different versions of SPM (SPM2, SPM5, and SPM8, although it hasn't been tested extensively with SPM2).
- Download spm_ss toolbox from http://www.nitrc.org/projects/spm_ss.
[See the readme.pdf file enclosed with the toolbox for details.]
1. Group-constrained subject-specific (GSS) analyses
Primary uses: (1) for developing new localizers; (2) for using in addition to the traditional random-effects (RFX) group analysis in cases where (i) the sample is too small, and/or (ii) high variability in the locations of activations is observed across subjects, and, consequently, the RFX analysis may not be sensitive enough.
[NB: Unlike what we did in the analyses reported in Fedorenko et al. (2010), where we used all but the first run to define the ROIs and extracted responses from the first run’s data, in this toolbox we implemented a more powerful way to use independent data subsets for ROI definition and response estimation: across-runs cross-validation. In particular, whenever the localizer contrast and the effect-of-interest contrast are not orthogonal, the toolbox creates additional contrasts: one for each run, and one for each set of runs orthogonal to each individual run. For example, for a four-run study, it creates additional contrasts for the following subsets of the data: for run 1, 2, 3, 4, and for runs 234, 134, 124, 123. It then uses the orthogonal contrasts (i.e., 234, 134, 124, and 123) from each subject to create the overlap map. Finally, it intersects the resulting partitions with each of the orthogonal contrast maps and extracts the response from the corresponding orthogonal run (e.g., intersecting with the map for runs 234 and extracting from run 1, intersecting with the map for runs 134 and extracting from run 2, etc.). The responses are then averaged across runs for each subject, and then across subjects.]
2. GSS fROI analyses with pre-existing parcels (using manually specified ROIs, mROIs)
Primary use: for defining subject-specific fROIs using parcels derived from some localizer task (or using anatomical parcels; e.g., Brodmann areas), and for extracting the response from these fROIs. (For example, you can use our parcels for intesecting with individual subjects’ activation maps for the S>N localizer contrast.)