基于样本熵与决策树的麻醉意识深度评价指数的研究  

Study on the Evaluation Index of Depth of Anesthesia Awareness Based on Sample Entropy and Decision Tree

作　　者：刘军[1] 周雅琪[1] 陈绍宾[2] 徐天昊 陈枭 谢斐[1] 

机构地区：[1]浙江大学生物医学工程与仪器科学学院生物医学工程教育部重点实验室,杭州310027 [2]浙江普可医疗科技有限公司,杭州310007

出　　处：《生物医学工程学杂志》2015年第2期434-439,共6页Journal of Biomedical Engineering

基　　金：浙江省自然科学基金资助项目(LY13H180004);中央高校基本科研业务费专项资金资助项目

摘　　要：麻醉意识深度监测是临床中保证全身麻醉(全麻)手术顺利进行的关键手段之一,脑电图(EEG)作为检测大脑皮层活动的主要信号,是评价麻醉意识深度的重要工具。本文根据脑电信号随麻醉意识深度变化的趋势,提出结合脑电分析中的时域、频域及复杂度方法,采用决策树分类器与最小二乘拟合法计算麻醉深度指数(DOAI)。利用临床采集的40例丙泊酚全麻手术患者的脑电信号和麻醉专家对信号的分类、评分对此方案进行验证,实验结果与目前临床上广泛使用的BIS指数进行对比,结果显示DOAI与BIS指数的Pearson相关性可达0.89,从而证实此方案的可行性与准确性,为麻醉监护工作者提供了一种思路。Currently,monitoring system of awareness of the depth of anesthesia has been more and more widely used in clinical practices.The intelligent evaluation algorithm is the key technology of this type of equipment.On the basis of studies about changes of electroencephalography（EEG）features during anesthesia,a discussion about how to select reasonable EEG parameters and classification algorithm to monitor the depth of anesthesia has taken place.A scheme which combines time domain analysis,frequency domain analysis and the variability of EEG and decision tree as classifier and least squares to compute Depth of anesthesia Index（DOAI）is proposed in this paper.Using the EEG of 40 patients who underwent general anesthesia with propofol,and the classification and the score of the EEG annotated by anesthesiologist,we verified this scheme with experiments.Classification and scoring was based on a combination of modified observer assessment of alertness/sedation（MOAA/S）,and the changes of EEG parameters of patients during anesthesia.Then we used the BIS index to testify the validation of the DOAI.Results showed that Pearson′s correlation coefficient between the DOAI and the BIS over the test set was 0.89.It is demonstrated that the method is feasible and has good accuracy.

关 键 词：麻醉监护 脑电参数 决策树 麻醉深度指数 BIS指数 

分 类 号：R614[医药卫生—麻醉学]