A Genetic Algorithm Approach for Localization of Deep Sources in MEG

The development of efficient methods for the reconstruction of sources underlying MEG measurements coming from deep brain activity is an important goal in neuroimaging research. In this communication we describe an optimization method based on a genetic algorithm (GA) of the multi-object type, to address the inverse problem, lending some weight to the proximity of the source to a given deep brain structure. The fundamental mechanism, as usual, operates on a population of individuals, each representing a current dipole or a set of dipoles (position and orientation), which is supposed to be a solution of the inverse problem and represented by binary coded strings of 51 bits for each source. The results show the efficiency of the algorithm to deal with neuromagnetic activity ascribed to deep brain regions.