Performance of a novel SQUID-based superconducting imaging-surface magnetoencephalography system

Performance for a recently completed whole-head magnetoencephalography system using a superconducting imaging surface (SIS) surrounding an array of 150 SQUID magnetometers is reported. The helmet-like SIS is hemispherical in shape with a brim. Conceptually, the SIS images nearby Sources onto the SQUIDs while shielding sensors from distant "noise" sources. A finite element method (FEM) description using the as-built geometry was developed to describe the SIS effect on source fields by imposing B, (surface) = 0. Sensors consist of 8 x 8 mm(2) SQUID magnetometers with 0.84 nT/Phi, sensitivity and <3 fT/rootHz noise. A series of phantom experiments to verify system efficacy have been completed, Simple dry-wire phantoms were used to eliminate model dependence from our results. Phantom coils were distributed throughout the volume encompassed by the array with a variety of orientations, Each phantom coil was precisely machined and located to better than 25 mum and 10 mRad accuracy. Excellent agreement between model-calculated and measured magnetic field distributions of all phantom coil positions, and orientations was found. Good agreement was found between modeled and measured shielding of the SQUIDs from sources external to the array showing significant frequency-independent shielding. Phantom localization precision was better than 0.5 mm at all locations with a mean of better than 0.3 mm. (C) 2001 Published by Elsevier Science B.V.