Anesthesia depth-dependent features of electroencephalographic bicoherence spectrum during sevoflurane anesthesia.

Abstract

Growth pattern in the electroencephalographic bicoherence spectrum has recently been found to relate to anesthetic depth, and bicoherence analysis can reflect behavior of the thalamocortical reverberating network. Because the thalamocortical network is known to represent a key factor in sleep by anesthesia, systematic and qualitative bicoherence studies of different anesthetic depths is necessary throughout all pairs of frequencies. Sixteen patients were anesthetized using sevoflurane (1, 2, or 3%) combined with remifentanil (0.4 microg x kg x min). Raw electroencephalographic signals were collected, and bicoherence was estimated in all pairs of frequencies, between 0.5 and 40 Hz at 0.5-Hz intervals. Sevoflurane (1%) caused two main peaks, spindle frequencies (11.0 +/- 1.2 Hz, 44.7 +/- 12.3% [bicoherence growth]) and delta-theta frequencies (5.4 +/- 0.5 Hz, 33.0 +/- 8.4%), in the diagonal line of biphasic bicoherence plots. High concentrations of sevoflurane (2% and 3%) shifted these peaks to 9.8 +/- 1.1 Hz, 46.2 +/- 12.7%; 8.7 +/- 1.3 Hz, 37.2 +/- 13.7% and 4.9 +/- 0.5 Hz, 44.6 +/- 7.0%; 4.3 +/- 0.8 Hz, 45.2 +/- 10.6%, respectively. Sevoflurane caused a third bicoherence peak to appear in another heterogeneous pair frequency (pair of alpha basal frequency and its double frequency), outside the diagonal line, which also inherited the behavior of alpha bicoherence peaks at different anesthetic depths. Sevoflurane anesthesia caused bicoherence peaks in alpha and delta-theta areas and also formed secondary third peaks. Deeper sevoflurane anesthesia shifted all bicoherence peaks to lower frequencies and caused increased bicoherence growth in the delta-theta area. The obtained features are consistent with characteristics of the thalamocortical reverberating network and suggest the importance of bicoherence analysis for the thalamic system.