|
/vol08/sleigh02/ | © Copyright 2001 | |||
| Volume 08 | Received: Accepted: |
15 Jul 2000 01 Nov 2000 |
|||
|
Subcortical vs cortical actions of general anaesthetics: A cellular automaton model
Sleigh, J. W., Steyn-Ross, M. L. & Steyn-Ross, D. A. |
|
| Abstract | |
| Cellular automaton models may be used to simulate cortical dynamics. In this study we investigated the effects of different sub-cortical modulation patterns and strengths on the observed "pseudoEEG" output from the array. Computer simulations were run with input strengths varying from 1 to 50 (50 = 3.125% of the array) cells per input step, and varying the patterns to be either regular or bursting. The main results were: (1) The spectral power is increased appreciably by increasing strength of the input. (2) The spectral power diverges near the critical-point of the array (signalling the phase-transition). This divergence was more marked with stronger input. (3) The power law scaling (1/f slope) of the pseudoEEG decreases abruptly near the critical-point - again indicative of the phase-transition - particularly when the input was weak. (4) The 50% spectral edge frequency (SE50) decreased from the awake to the anaesthetized state. The decrease in frequency of the first time-derivative (~differenced, called the SE50d) pseudoEEG signal was the output parameter that was most independent of the input magnitude. It may be concluded that: (1) A combination of SE50d, 1/f slope, and spectral power divergence may most reliably detect loss, and regain, of consciousness in the clinical setting. (2) The dependence of spectral power on variations in ('subcortical') input strength may explain some of the failure of EEG-derived parameters in the clinical detection of level of consciousness between different patients, and differing anaesthetic regimens. | |
| Full Text |
|
|