Effects of Measurement Errors and Noise on MEG Moving Dipole Inverse Solutions

Magnetoencephalograms (MEG's) are increasingly being used with the moving dipole method to localize electrical sources in the brain. In this method, also known as the dipole location method, a dipolar source is moved about in a model of the head while its amplitude and orientation are also adjusted to obtain a solution dipole which gives the least squares error fit between the measured MEG's and those produced by the dipolar source. The accuracy of this solution is affected by various measurement errors such as errors in the size of the measurement grid, size of the head model, etc., and by noise in the measured MEG's. This study uses computer modeling methods to investigate the effects of these factors on the localization accuracy of sources in the cortical region of the brain for several different ways of making MEG measurements using single channel and/or multichannel detectors.