Clinical implementation of advanced control in anaesthesia

Abstract

A multi-sensor fusion system for monitoring and control of depth of anaesthesia in the operating theatre is described. Two types of anaesthesia measures are considered, the first being based on auditory evoked responses, while the second measure is based on cardiovascular system signals. The first measure is based on analysis of the signal using wavelets and a neuro-fuzzy system, while the second measure is based on analysis of the measurements using a rule-based fuzzy logic system. The two measures are combined to give a final level of depth of anaesthesia using another rule-based fuzzy logic system. The depth of anaesthesia level is fed back to a target controlled infusion system (TCI) to control the infusion of the drug propofol for maintaining a constant level of depth of anaesthesia.

Clinical Decision Support and Closed‐Loop Control for Intensive Care Unit Sedation

“…The same system was used in , with an auditory evoked potential as the control variable. Intravenous propofol anesthesia has also been delivered by a closed‐loop controller that uses both auditory evoked responses and cardiovascular responses as the control variables with a fuzzy‐logic algorithm in . This system has had only very minimal clinical testing.…”

Detection and material identification of loose particles inside the aerospace power supply via stochastic resonance and LVQ network

“…Both time-domain and frequency-domain methods are extensively used for eigenvector extraction. The linear prediction coding, a typical time-domain method, analysing the correlation of sampling data, was adopted to classify the signal of loose particles within relays (Wang et al, 2007) but also processes high-frequency wavelet coefficients subspace (Deng and Ling, 1999;Carnero and Drygajlo, 1999;Linkens et al, 2000;Philippe et al, 1999), suitable for processing a particle signal with wide energy distribution. To obtain more material information about the particle signal, wavelet packet transform instead of Fourier transform is employed to extract eigenvectors in this paper.…”

Adaptive control for non‐negative and compartmental dynamical systems with applications to general anesthesia

“…Propofol is an intravenous anesthetic that has been used for both induction and maintenance of general anesthesia [32]. A simple yet effective patient model for the disposition of propofol is based on the three-compartment mammillary model shown in Figure 1 with the first compartment acting as the central compartment and the remaining two compartments exchanging with the central compartment [33,34]. The three-compartment mammillary system provides a pharmacokinetic model for a patient describing the distribution of propofol into the central compartment (identified with the intravascular blood volume as well as highly perfused organs) and the other various tissue groups of the body.…”

“…where V c is the volume in liters of the central compartment. As noted in [34], V c can be approximately calculated by V c =(0.159 '/kg)(M kg), where M is the weight (mass) in kilograms of the patient. In our control design we assume M=70 kg so that the desired level of propofol mass in the central component is given by x d1 =(4 mg/m')(0.159'/kg)(70 kg)=44.52 mg.…”