基于最大似然法的中风病常用中成药适宜人群经典判别分析
2018-09-18罗文舒胡湘杨卓欣于海波皮敏张金文黄美玲崔梦媛
罗文舒 胡湘 杨卓欣 于海波 皮敏 张金文 黄美玲 崔梦媛
【摘 要】 目的:应用基于最大似然法的经典判别分析研究中风病常用中成药适宜人群的临床特征。方法:查询电子处方集系统,筛选应用量最大的前3种中成药,利用经典判别分析对常用中成药适宜人群建立数学模型,分析模型对不同人群的预测能力。结果:筛选出中风病应用量最大的前3种中成药为灯盏生脉胶囊、银杏叶滴丸、强力天麻杜仲胶囊,共纳入203例受试者,利用经典判别分析构建数学模型,对3种中成药组测试集进行预测,总判别正确率为76.24%。结论:利用基于最大似然法的经典判别分析构建中成药适宜人群的数学模型,对3种中成药适宜人群有良好的识别效果。
【关键词】 中风病;中成药适宜人群;经典判别分析
【中图分类号】R255.2 【文献标志码】 A【文章编号】1007-8517(2018)15-0010-03
Abstract:Objective To investigate the recognition of proper population of Chinese patent medicine commonly used in stroke by classical discriminant analysis. Methods The first 3 Chinese patent medicines with the largest number of applications have been screened by inquiring of electronic medical records. The classical discriminant analysis based on Maximum Likelihood is performed to construct the mathematic model for the proper population of Chinese patent medicine commonly used in stroke. Results The first 3 Chinese patent medicines with the largest number of applications are Erigeron Shengmai Capsule, Ginkgo biloba drop pills and Powerful Gastrodia Eucommia Capsule. The mathematic model is constructed by classical discriminant analysis for 203 subjects. Total recognition accuracy for 3 groups of Chinese patent medicines is 76.24%. Conclusions The recognition accuracy of the mathematic model for proper population of Chinese patent medicine constructed by classical discriminant analysis is satisfied.
Keywords:Stroke; Proper Population of Chinese Patent Medicine; Classical Discriminant Analysis
中风病是严重危害我国人民健康的重大疾病,其发病率、致死率和致残率均很高,近年脑卒中发病率和患病率仍在持续上升[1]。中成药对脑血管疾病疗效好,毒副作用少,已成为我国中风病患者治疗的重要选择之一,是临床防治中风病的重要药物[2]。随着中成药在临床上的广泛应用,其药源性疾病发生有增多趋势[3]。临床上很多医师特别是西医师,因对中成药药理理解不充分,使用时缺少辨证论治,造成了不合理使用现象[4],这已引起行业内重視[5]。因此,构建符合中医理论的中风病常用中成药适宜人群的数学模型,为临床上提供参考,有助于提高疗效,控制成本,促进康复。……
