The use of fuzzy integrals and bispectral analysis of the electroencephalogram to predict movement under anesthesia

摘要: The objective of this study was to design and evaluate a methodology for estimating the depth anesthesia in canine model that integrates electroencephalogram (EEG)-derived autoregressive (AR) parameters, hemodynamic alveolar anesthetic concentration. Using parametric approach, two separate AR models order ten were derived EEG, one from third-order cumulant sequence other autocorrelation lags EEG. Since dose versus curve is highly nonlinear, neural network (NN) chosen as basic estimator multiple NN approach conceived which took EEG concentration input feature vectors. estimation involves cognitive well statistical uncertainties, fuzzy integral used integrate individual estimates various networks arrive at final estimate anesthesia. Data 11 experiments train NN's then tested on nine experiments. (when 43 vectors seven test experiments) classified 40 (93%) them correctly, offering substantial improvement over estimates.