Somatosensory evoked potential correlates of psychophysical magnitude estimations for tactile air-puff stimulation in man.

Brief air-puff stimuli were applied to the volar surface of the right hand to obtain both psychophysical and neurophysiological responses. The detection threshold (So) was first determined (0.56 kg . cm-2 +/- 0.20 kg . cm-2, mean +/- SD) and six levels of the stimulus intensities (So + 0.25 kg . cm-2, So + 1.25 kg . cm-2, So + 2.50 kg . cm-2, So + 3.75 kg . cm-2, So + 5.00 kg . cm-2, and So + 6.25 kg . cm-2) were employed for magnitude estimation using the stimulus level of So + 2.50 kg . cm-2 as the standard stimulus. The subject was asked to estimate numerically the series of stimulus intensities randomly presented. Cortical SEPs were recorded over the hand sensory area in response to a set of 120 air-puffs at the identical intensity level. Thus SEPs for six sets of stimulus intensities given in a random order were obtained from each subject. Six components (N20, P27, N35, P45, N60, and P75) were recorded within 100 ms following stimulation. It was seen that a simple power function with an exponent of 0.81 could be an adequate description of the stimulus-response function for magnitude estimation, as was also revealed by the high correlation coefficient (r = 0.98). Similarly, stimulus-amplitude functions of different SEP components were well represented by straight lines in double logarithmic plots. The function of the early P27-N35 had the highest exponent (0.56) and also the highest correlation coefficient (r = 0.91). Plotting subjective magnitude of the abscissa produced power functions similar to the stimulus-amplitude functions. However, higher correlations were observed for later components. The amplitudes of the four earlier components correlated with stimulus intensity when the effect of subjective magnitudes was removed. In contrast, the correlation between amplitudes and subjective magnitudes with stimulus intensity held constant was positive and significant for the later three components. These results may indicate that early SEP components represent neural coding of physical intensity while later components are more closely related to the subjective judgment of the stimulus.