Multichannel recording of median nerve somatosensory evoked potentials

Clinical applications of multichannel (>or=64 electrodes) electroencephalography (EEG) have been limited so far. Amplitude variability of evoked potentials in healthy subjects is large, which limits their diagnostic applicability. This amplitude variability may be partially due to spatial undersampling of anatomical variations in cortical generators. In the present study, we therefore investigated whether 128-channel recordings of somatosensory evoked potentials (SEPs) can reduce this amplitude variability in healthy subjects. Additionally, we explored the relation between amplitude and age.

We recorded median nerve SEPs using a 128-channel EEG system in 50 healthy subjects (20-70 years) and compared N20, P27, and P45 amplitude as obtained with a 128-channel analysis method - based on butterfly plots and spatial topographies - and as obtained using a conventional one-cortical-channel configuration and analysis. Scalp and earlobe references were compared.

Although amplitude variability itself was not reduced, a reduced coefficient of variation was obtained with the 128-channel method due to higher SEP amplitudes, compared to the conventional one-channel method, independent of reference.

These results suggest that at the cost of some additional preparation time, the 128-channel method can measure SEP amplitude more accurately and might therefore be more sensitive to physiological and pathological changes. For optimal amplitude estimation, we recommend to increase the number of centroparietal electrodes or, preferably, to perform at least a 64-channel recording.