Baseline optimization for noise cancellation systems [of SQUID MEG]

SQUID magnetoencephalography systems (MEG) are usually subjected to unacceptable levels of environmental noise even when operated within shielded rooms. The cancellation of this noise is accomplished by configuring the SQUID sensors either as hardware or synthetic gradiometers or adaptive systems. The baseline length of such systems not only lowers the environmental noise but it may also reduce signals detected from the brain. The resulting signal-to-noise ratio is a function of baseline and it peaks at the optimum baseline length. The optimum baselines were investigated by numerical computations and also experimentally for four existing MEG installations. In all four cases it was found that the optimal baselines are quite short (less than 8 cm)