Somatosensory responses to stimulus trains: normative data

Abstract

This paper investigates the human evoked response produced by peripherally stimulating the median nerve with a repetitive high frequency stimulus. The human somatosensory system can be driven and the resultant evoked response can be consistently recognized at frequencies as high as 200 Hz. This somatosensory evoked response to trains (SERT) has proven to be a very reliable and easily quantifiable measure. Seventeen normal subjects and a total of 28 hands were investigated. Each hand was stimulated with eight different frequency runs, each consisting of 64 trains of stimuli spaced 2 sec apart. The frequencies of stimulation were 12, 20,40,60,80,100,160 and 200 Hz. Each stimulus had a width of 0.25 msec and the train lenght was fixed at 250 msec. Three data processing steps were utilized. The initial procedure was the on-line computation of the contralateral SERT. Secondly, a filtered SERT was calculated by narrowband filtering of the EEG signal and averaging. Finally, a single cycle filtered SERT was calculated by narrowband filtering of the EEG signal and averaging across the interstimulus interval. The amplitude and time of occurence of the peak negative value obtained from this measure were used to quantify the suject response at each stimulating frequency. Signal variability due to activity at the bipolar reference electrode was investigated, as were differential phase shifts between signals recorded at anterior and posterior electrode locations. In addition, topographical mappings of the single cycle amplitudes and phase angles were obtained at the different stimulating frequencies. The distributions of the peak-to-peak amplitude values were consistent with topographical data previously reported for the somatosensory evoked response, while the phase angles tended to distribute in such a way as to divide the scalp into two discrete regions.