1、基于基于MatlabMatlab智能算法的胃食管返流病智能算法的胃食管返流病中药治疗方案优选中药治疗方案优选About GERDAbout GERDHigh Incidence Rate(5.2%-8.5%in east Asia after 2005)Three sub typesAcid reflux and heartburnLoss of control by PPIsHerbs are a good alternative choiceGood to be evaluated by symptom scores,which is somehow like TCM evaluation
2、Possible to be prescribed by machine based on studyANN is an important way to realize thatRough set used for the diagnosis of GERDRough set used for the diagnosis of GERDAfter the diagnosis,then?Where is the corresponding treatments?What is ANN?Artificial neural networks(ANNs),a form of connectionis
3、m,are computing systems inspired by the biological neural networks that constitute animal brains.Such systems learn(progressively improve performance)to do tasks by considering examples,generally without task-specific programming.For example,in image recognition,they might learn to identify images t
4、hat contain cats by analyzing example images that have been manually labeled as cat or no cat and using the analytic results to identify cats in other images.They have found most use in applications difficult to express in a traditional computer algorithm using rule-based programming.Why we need?1.A
5、I challenges increased in human society,including those traditional fields;2.The idea of TCM has big possibility to be simulated by ANN.3.The development of some toolbox make it easier to make research into this field.4.Matlab is a good tool for Simulink and explore the possibility of“Treatment base
6、d on symptoms recognition”.Human Training and Machine Study 1.Different Study curve 2.Different Capacity of study 3.Different precision of acquiring knowledge 4.Different utilization powera3=f3(LW3,2f2(LW2,1f1(IW1,1pb1)b2)b3)=y Correspondance Bwteen Symptoms and Herbs D1D2DnS1 S1 Sm Frequency normal
7、ized0,1（xmin）/(maxmin)Memory Weight MatrixX=0000.00000.01364.00909.00000.00455.00000.00000.00000.00000.00455.01364.00000.00455.00909.07727.01364.00455.00000.00000.00455.00455.02273.02273.00000.00000.01818.00909.02273.03182.00000.00000.02727.00909.00000.03182.00455.00000.01818.02273.00455.01818.00000
8、.00455.00909.00000.00000.00909.00455.00455.00000.00909.00909.00455.00000.00455.00455.00909.00455.03636.00000.00455.02273.01364.00000.02273.00000.00455.00909.00000.01364.03182.00455.00000.02273.02273.01818.03636.00000.01364.05909.03636.02727.03636.00455.00455.04545.01818.01364.04545.00000.00455.04545
9、.01818.00909.03182.00455.00455.02727.01364.00909.04545.01364.00455.03636.01818.00909.04091.00000.00000.00909.00455.02273.01818.00000.00000.03636.01818.00000.01818.00909.00455.03182.00909.00909.00909.00000.00455.00909.00455.00909.03636.00455.00000.03636.03182.02273.02727.00455.01364.05455.01818.02273
10、.03636.00000.00000.05000.03636.00455.04545.00000.00000.03636.01364.02273.02727.00909.00909.06818.03636.02273.06818.01818.00909.04545.02727.01364.03636.00909.00909.00000.15909.01818.05455.00455.00909.05455.02273.02273.05909.0 Memory weight matrixS1:0.3636;S2:0.2273;S3:0.2182;S4:0.5272;S5:0.0909;S6:0.
11、0455;S7:0.4545;S8:0.2273;S9:0.3636;S10:0.4909;S11:0.0455;S12:0.0818。Lower threshold of selectionSymptoms and Herbs Sets Say s1=中酸，s2=嘈杂，s3=反胃，s4=呕吐，s5=胸痹，s6=梅核气，s7=胸痞，s8=呕逆，s9=胃痞，s10=纳差，s11=胁痛，s12=嗳气。药物d1=陈皮，d2=甘草，d3=白术，d4=半夏，d5=苍术，d6=茯苓，d7=人参，d8=生姜，d9=神曲，d10=吴萸，d11=砂仁，d12=黄连。Composition of any four
12、 symptoms:MatLab program:combntns(1,2,3,4,5,6,7,8,9,10,11,12,4)S1=1 1 1 1 0 0 0 0 0 0 0 0;S2=1 1 1 0 0 1 0 0 0 0 0 0;S3=1 1 1 0 0 0 0 1 0 0 0 0;D1=1 1 1 1 1 1 0 1 0 1 0 1;D2=1 1 1 1 1 1 0 1 0 1 0 1;D3=1 1 1 1 1 1 0 1 0 1 0 1;S=S1;S2;S3;D=D1;D2;D3;net=newff(0,1;0,1;0,1;0,1;0,1;0,1;0,1;0,1;0,1;0,1;0,1
13、;0,1,5,3,12,tansig,tansig,tansig);Symptoms setBP networkTraining and Simulink Train function for training and Sim function for Simulink:After around 3090 paces,the error is almost approaching 0Case Test 1 The patient has反酸，胃痞，纳差，嗳气symptoms，so the symptoms set would be：S4=1 0 0 0 0 0 1 0 1 0 0 1;D4=s
14、im(net,S4)结果显示应该采用的药物为陈皮，半夏，苍术，黄连，基本勾勒出了苍连丸（见古今医鉴卷五）的主要成份。苍连丸由苍术、陈皮、半夏、黄连、茯苓、吴萸六味药组成，用治郁积吞酸。患者在反酸的同时症见胃痞、纳差、嗳气，恰为郁积的表现，因此基本符合苍连丸的应用指征。Case Test 2 Patient 2 has嘈杂，胸痹，胁痛，嗳气symptoms，the symptoms set would be:S5=0 1 0 0 1 0 0 0 0 0 1 1;D5=sim(net,S5)据其输出结果，可选用的药物为陈皮，甘草，白术，半夏，苍术，茯苓，生姜，吴萸，黄连，基本包含了二陈、左金等对治
15、之剂，能够较好地体现中医的辨证论治思想。上述过程的输出层也可应用硬限幅函数“hardlim”，则结果将以“0”或“1”的二值形式出现，0表示该药不选用，1表示选用该药。Programed Realization of TCM D&T对于一种疾病，首先要通过大量学习样本的研究，得出哪些为该病的主要症状及证型，然后通过Rough集分析，进行属性集的约简、属性值的约简，导出决策规则，这就是“辨证”的过程；在此基础上，广泛学习古今各种文献，建立数据库，并通过前述“记忆权数”矩阵，学习相关“论治”规则，通过神经网络的训练与泛化，建立ANN模型，这就是“论治”的过程。然后编写程式，通过人机交互界面，完成机器“辨证论治”的自动过程，这也许是新一代中医专家系统的一个理想模式。
