Spatial resolution of neuromagnetic records: theoretical calculations in a spherical model.

Recently developed small arrays of SQUID- (superconducting quantum interference device) based magnetic sensors can, if appropriately placed, locate the position of a confined biomagnetic source without moving the array. A technique is presented having a relative accuracy of about 2% for calibrating such sensors having detection coils with the geometry of a second-order gradiometer. The effects of calibration error and magnetic noise on the accuracy of locating an equivalent current dipole source in the human brain are investigated for five- and seven-sensor probes and for a pair of seven-sensor probes. With a noise level of 5% of peak signal, uncertainties of about 20% in source strength and depth for a five-sensor probe are reduced to 8% for a pair of seven-sensor probes, and uncertainties of about 15 mm in lateral position are reduced to 1 mm for the configuration considered. >

SQUID arrays for simultaneous magnetic measurements: calibration and source localization performance

: This final report includes descriptions of substantive experimental studies of neural phenomena related to attention and auditory perception. It also describes efforts to enhance the superconducting instruments and other devices needed for the rapid and accurate accumulation of neuromagnetic data, and advances made in techniques for calibrating these instruments and for analyzing neuromagnetic data. The substantive experiments included a major study of the magnetic N100 phenomena and its sources and how they are affected by selective attention. Its relationship to the electrical N100 is considered, and required future research described. Also, work on the magnetic P300 phenomenon is described. This work confirmed earlier studies showing that the equivalent current dipole source is located in or near the hippocampal formation. The localization of multiple auditory sources is described. Improvements in instrumentation include the installation of a new gantry for purposes of evaluation, the design of a novel device for quantifying positions in magnetic resonance images, and the development of a graphics program for depicting a current dipole in the heads of subjects are also described. New methods for calibrating multisensor systems were developed, and the details are provided in the report. Finally, an opportunity arose during the course of this project to located a very small metallic object accidentally embedded in the back of a human patient. This allowed us to obtain surgical verification of magnetic methods for locating sources. The predicted position was accurate within two millimeters. Keywords: Selective attention, Neural activity, Magnetic localization.

M. J. Kajola

Papers

Design of a 7-channel DC-SQUID gradiometer for brain research