Interactions of distant sources of brain magnetic fields: estimation errors of single-dipole model

The simplest and most widely used method in the biomagnetic inverse problem is the nonlinear parameter estimation using single-dipole model. However, when the number of activated brain areas is more than one, this simplest dipole solution leads to estimation errors. We evaluated the errors of position, strength, and orientation of the sources estimated by using this single-dipole method when applied to twin-dipole activity using locally-selected channels rather than all channels of the whole-head measurement system. The results suggest the errors generated by the single-dipole method using locally selected channels may be acceptable in some conditions, such as estimating activities of both auditory cortices.