Neural source reconstruction of MEG/EEG data requires subject specific geometrical data (i.e. electrode locations and anatomical brain data). For many reconstruction techniques, results are volumetric rather than surface-based. A group level analysis for volumetric data has not yet been implemented in MNE Python. The proposed project aims to fill this gap, by implementing necessary tools for group level analyses based on volumetric data, among which are: non-linear warping of one volumetric (grid) space to another; creating pseudo-individual anatomical MR images, based on a subject’s head shape; and output preparation, such that it can be used with already built in statistical functions. The results will be a set of Python functions that enable the user to prepare individual volumetric subject data for group level analyses. Furthermore the respective visualization will be targeted as well.


Tommy Clausner


  • Eric Larson
  • Jean-Rémi King
  • Mainak Jas
  • Denis Engemann
  • Alex Gramfort