THE CONTRIBUTION OF AFFERENT FIBER TRACTS TO THE SOMATOSENSORY EVOKED POTENTIAL *

The short-latency components of the somatosensory evoked potential (SLSEP) provide an index of the integrity of the neuraxis from peripheral nerve to cortex. The full clinical utility of this technique is dependent on the degree to which each component can be ascribed to an anatomical structure and related to a portion of the ongoing neuroelectric activity in this region. Identification of the intracranial and intraspinal sources of the surface SLSEP components is a formidable task, which requires a three-dimensional analysis of the field potential distributions and the simultaneously recorded multiple unit activity. For a description of the techniques of source identification see the review by Vaughan in this volume. The results of our mapping studies in the monkey, following both median and peroneal nerve stimulation have been previously reported14 and the following is a brief summary and discussion of these results. Using a noncephalic reference, a series of five small positive components, which precede the initial cortical activity, were identified overlying the contralateral sensorimotor cortex. The latencies of the components are dependent on the length of the conduction pathways and varied with the stimulated nerve, point of stimulation and size of the animal; however, the response waveform was similar following either hindlimb or forelimb stimulation. On the basis of timing and topography the source of SLSEPl was localized to the proximal portion of the peripheral nerve, including the dorsal root fibers. The dorsal roots are angled relative to the ascending spinal tracts and SLSEPl is therefore seen as a “near-field’’ component overlying the spinal zone of root entry and as a “far-field” potential overlying rostral spinal and brainstem sites. At the vertex this potential is small and detected only as an inflection on the rising edge of SLSEP2. SLSEP2 is recorded as a traveling wave overlying the spinal cord rostral to the segments of root entry and as a fixed latency component above the level of the dorsal column nuclei. This component reflects propagated activity of the most rapidly conducting axons within the ascending spinal tracts, principally the dorsal columns. The administration of the neurotoxin acrylamide, which initially damages the preterminal axons in the gracile nucleus, distorts SLSEP2 following peroneal nerve stimulation, but leaves the response to median nerve stimulation unaffected. SLSEP3 and SLSEP4 are recorded as large potentials in the ventral thalamus, but each of these components can be traced with decreasing latency to caudal sites in the brainstem. The topography of the components suggest that they represent sequential activity in the mesencephalic and diencephalic portions of the medial lemniscus. SLSEPS can be traced a t increasing latencies from the thalamus to the sensorimotor