Pain perception, obstructive imagery and phase-ordered gamma oscillations

Abstract

The neural mechanisms underlying pain perception and anti-nociceptive effects of mental imagery are not well understood. Using a measure of phase-ordered beta and gamma EEG oscillations in response to painful electric stimulation, we recently found that somatosensory event-related phase-ordered gamma oscillations (38–42 Hz), elicited by the onset of painful stimuli over Cz scalp site, were linearly related to pain perception. In the present study, 38 subjects were engaged in a painful stimulus detection task using an oddball paradigm. This task was performed under a condition in which subjects were required simply to count the number of target stimuli (pain condition) and under another condition in which subjects were required to produce an obstructive mental imagery of painful stimulus perception (obstructive imagery). Only EEG responses to standard stimuli were analyzed in this study. Correlation analysis of sweeps for each individual revealed brief intervals of phase ordering of EEG patterns in the beta and gamma bands. The frequencies of interest were the beta1 (26–30 Hz), beta2 (30–34 Hz), gamma1 (34–38 Hz), gamma2 (38–42 Hz) and gamma3 (42–46 Hz) bands. Obstructive imagery treatment, compared to pain condition, significantly reduced pain perception. This reduction was paralleled by significant decreases of evoked phase-ordered gamma2 and gamma3 patterns over Cz scalp site. Phase-ordered oscillations at Cz scalp site, for both gamma2 and gamma3 bands, significantly predicted pain ratings during pain condition. Phase-ordered oscillation scores, obtained for these gamma bands over parietal and frontal scalp sites, resulted the best predictor of pain ratings during obstructive imagery. This study provides evidence for the role of gamma oscillations in the subjective experience of pain. Further, it has provided support for the view that pain reduction during obstructive mental imagery is the product of an inhibitory process involving frontal and parietal cortical regions.