Intersession replicability of dipole parameters from three components of the auditory evoked magnetic field.

The replicability of dipole localizations between sessions in an unselected group of subjects was studied. Auditory evoked magnetic fields (AEMFs) in response to contralaterally and ipsilaterally presented 1 kHz tone bursts were recorded from the right hemisphere of 12 subjects with normal hearing in two replicate sessions several days apart. Three long-latency components of the AEMF were studied, occurring at latencies near 50 msec (P1m), near 100 msec (N1m) and near 165 msec (P2m). A spherical model of the head was used to fit equivalent-current dipoles to the data. Statistical analysis of dipole parameters revealed virtually no differences between the two testing sessions. The variability between sessions had a mean absolute difference of 3 to 10 mm for the spatial parameters. Comparison of dipole parameters between components showed that there was a replicable, but nonsignificant, trend for a difference in the location of the N1m from contralateral vs. ipsilateral stimulation, and a statistically significant confirmation that the P2m is located anterior to the N1m for contralateral stimulation. Magnetic resonance images from each subject were used to locate the dipoles near the primary auditory cortex in the Sylvian fissure.