A method for determining optimal interelectrode spacing for cerebral topographic mapping

Abstract

Inappropriately large interelectrode spacing may result in significant errors of localization and erroneous measurements of amplitudes, due to aliasing. Oversampling with very large numbers of scalp electrodes is impractical. We present a generally applicable method for determination of optimal interelectrode spacing. This optimal spacing is computed from the Nyquist distance for the potential of interest. For most potentials, the bandwidth of the topographic distribution is not known, so it must be first measured by oversampling with tightly spaced electrode arrays. The Nyquist distance will vary for different types of cerebral potentials, different components of a given potential, and for different areas of the scalp. We have determined the Nyquist distances for interelectrode spacing necessary to record accurately the geographically small components of the SEP in the region of the sensory strip. These distances are less than 3 cm, which is significantly smaller than the 7 cm distance of the 10-20 system. Recordings of SEPs in this region made with larger spacings may have significant errors. Recordings made at the Nyquist distance may be precisely interpolated to arbitrary accuracy.