Cortical representation of the human hand assessed by two levels of high-resolution EEG recordings.

Increasing interest in cortical plasticity has prompted the growing use of somatosensory evoked potentials (SEPs) to estimate changes in the cortical representation of body regions. Here, we tested the effect of different sites of hand stimulation and of the density of spatial sampling in the quality of estimation of somatosensory sources. Sources of two SEP components from the primary somatosensory cortex (N20/P20 and P45) were estimated using two levels of spatial sampling (64- vs. 128-channel) and stimulation of four distal sites in the upper limbs, including single digits (first vs. fifth) and distal nerves with comparable cortical projection (superficial branch of the radial nerve and distal ulnar nerve). The most robust separation of somatosensory sources was achieved by comparing the cortical representations of the first digit and the distal ulnar nerve territories on the N20/P20 component of SEPs. Although both the 64- and the 128-electrode montages correctly discriminated these two areas, only the 128-electrode montage was able to significantly separate sources in the other cases, notably when using first versus fifth digit stimulation. Trustworthy distinction of cortical representations was not obtainable when using the P45 component, probably because of greater activation volume, radial orientation of sources in areas 1-2 and increased variability with attention and vigilance. Assessment of tangential SEP components to stimulation of first digit versus ulnar nerve appears the best option to assess plastic somatosensory changes, especially when using relatively low-electrode sampling.