This toolbox provides support for surface-based information mapping of functional Magnetic Resonance Imaging (fMRI) data, which is a data-driven application of Multi-Voxel Pattern Analysis.

Compared to traditional information mapping, as introduced by Nikos Kriegeskorte and colleagues in 2006, this has the following advantages:

1. -Easier visualization of results, especially when using inflated surfaces.
   

2. -More statistical power after multiple comparison correction, as comparisons are restricted to grey matter.
   

3. -Better inter-subject alignment, as it is based on folding patterns of the cortex rather than the less flexible but typical affine transformation.
   

4. -Better and more valid spatial specificity by the use of a geodesic distance metric, where distances are measured along the folded cortical surface.
   

5. -Improved discriminability between conditions of interest, as only grey matter is selected for multivariate analyses and thus less uninformative voxels are used.
   

The source code is written in Matlab and can be used with any neuroimaging analysis software package (e.g. AFNI, SPM, FSL). To get started, the documentation gives a detailed overview of the core functions of the toolbox. It also provides step-by-step instructions for analyzing data using the toolbox with an example dataset. An example with CoSMoMVPA is provided as well. Finally, the  documentation answers Frequently Asked and Anticipated Questions.

For more details, please visit the project pages on sourceforge or github. Feel free to contact us at p@q or x@y,

where p=nikolaas.oosterhof, q=unitn.it, x=tobias.wiestler.09, y=alumni.ucl.ac.uk.