PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Focal and rotationally invariant source reconstruction for EEG/MEG
Stefan Haufe, Vadim Nikulin, Klaus-Robert Müller, Andreas Ziehe and Guido Nolte
In: Biosignalverarbeitung 2008, Innovationen bei der Erfassung und Analyse bioelektrischer und biomagnetischer Signale, 16-18 Jul 2008, Potsdam, Germany.


Simultaneous localization of close neuronal sources from EEG/MEG measurements is a challenging problem. To solve it, we here propose Focal Vector Field Reconstruction (FVR) as a new source imaging technique. The FVR objective facilitates at the same time sparsity and smoothness of the solutions while it is invariant with respect to rotations of the coordinate system. In the simultaneous localization of left and right somatosensory N20 generators from real EEG recordings, FVR outperforms state-of-the-art methods (such as LORETA and Minimum Current Estimate), as it is the only method that delivers correct location of the source in the somatosensory area of each hemisphere; in accordance with neurophysiological prior knowledge.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Talk)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
Brain Computer Interfaces
ID Code:4266
Deposited By:Stefan Haufe
Deposited On:07 February 2009