Topography of somatosensory evoked potentials after stimulation of the median nerve.

We studied topography of major negative-positive peaks, NI, PI, NII, PII and NIII, of scalp recorded somatosensory evoked potentials (SEP) after stimulation of the median nerve. Unlike the diffusely distributed P14, NI, PI and NII recorded from contralateral hemisphere after unilateral stimulation normally showed statistically significant latency increase from frontal (N17, P20, N29) to central (N19, P23, N32) and parietal (N20, P26, N34) electrodes. However, NIII (N60) had considerable inter- and intra-individual variations with no consistent antero-posterior latency shift. In contrast to well localized N19 and N32 peaks at the contralateral central electrode, the N17, P20 and N29 peaks were registered over the bifrontal and ipsilateral central regions as well as in the vertex. The parietal N20 peak was also present at the occipital electrodes bilaterally. In patients with localized cerebral lesions, types of SEP abnormalities varied considerably, presumably reflecting complex somatosensory afferent pathways. A small lesion in posterolateral thalamus may totally eliminate NI, NII and NIII components over both hemispheres, sparing only P14 whereas a sizable lesion in the frontal or parietal lobe may affect only NII or NIII. Capsular lesions spare P14 and frontal N17 but may alter all the subsequent SEP components, or NII or NIII selectively. In some cases, the corresponding peaks at the central and parietal electrodes may be affected independently. The complex relationships between the type of SEP abnormalities and the location of cerebral lesions can best be explained by postulating the presence of multiple, at least partially independent, thalamocortical projections mediating regionally specific somatosensory inputs.