Kern S E, Johnson J O, Westenskow D R
Department of Anesthesiology, University of Utah, Salt Lake City 84132, USA.
Artif Intell Med. 1997 Sep;11(1):9-31. doi: 10.1016/s0933-3657(97)00022-5.
In this paper, we investigate the ability of fuzzy to adapt the parameters of a pharmacokinetic and pharmacodynamic model-based controller for the delivery of the muscle relaxant pancuronium. The system uses the model to control the rate of drug delivery and uses feedback from a sensor which measures muscle relaxation level to adapt the model using fuzzy logic. The control strategy administers mini-bolus doses of pancuronium and modulates the magnitude and time interval between the bolus doses to maintain a patient's muscle relaxation within an allowable range specified by the user. Before each new dose is given, the fuzzy logic adaptation scheme uses the error between the predicted patient response and the measured response to adapt the model. The system was tested using computer simulation by varying the parameters of the model by 50% from their nominal values. It was also evaluated in a clinical trial of five patients undergoing surgical procedures lasting 5 h or longer.
在本文中，我们研究了模糊逻辑用于调整基于药代动力学和药效学模型的控制器参数的能力，该控制器用于肌肉松弛剂泮库溴铵的给药。该系统使用该模型来控制药物输送速率，并利用来自测量肌肉松弛水平的传感器的反馈，通过模糊逻辑来调整模型。控制策略给予泮库溴铵小剂量推注，并调节推注剂量之间的大小和时间间隔，以将患者的肌肉松弛维持在用户指定的允许范围内。在给予每一新剂量之前，模糊逻辑自适应方案利用预测的患者反应与测量反应之间的误差来调整模型。通过将模型参数从其标称值变化50%进行计算机模拟，对该系统进行了测试。还在5名接受持续5小时或更长时间外科手术的患者的临床试验中对其进行了评估。
