[1]王海燕,鲁思博,孟 军,等.基于人工神经网络技术辨识妇科疾病证素-证型逻辑关系[J].医学信息,2020,(11):1-4.[doi:10.3969/j.issn.1006-1959.2020.11.001]
 WANG Hai-yan,LU Si-bo,MENG Jun,et al.Identification of Gynecological Disease Syndrome-syndrome Logic Relationship Based on Artificial Neural Networks Technology[J].Medical Information,2020,(11):1-4.[doi:10.3969/j.issn.1006-1959.2020.11.001]
点击复制

基于人工神经网络技术辨识妇科疾病证素-证型逻辑关系()
分享到:

医学信息[ISSN:1006-1959/CN:61-1278/R]

卷:
期数:
2020年11期
页码:
1-4
栏目:
医学信息学
出版日期:
2020-06-01

文章信息/Info

Title:
Identification of Gynecological Disease Syndrome-syndrome Logic Relationship Based on Artificial Neural Networks Technology
文章编号:
1006-1959(2020)11-0001-04
作者:
王海燕鲁思博孟 军
(1.广东医科大学图书馆,广东 东莞 523808;2.阿卡迪亚大学计算机科学与技术,新斯科舍 沃尔夫威尔 4725;3.深圳南山人民医院妇科,广东 深圳 518057;4.广东医科大学药学院,广东 东莞 523808)
Author(s):
WANG Hai-yanLU Si-boMENG Junet al
(1.Library of Guangdong Medical University,Dongguan 523808,Guangdong,China;2.Computer Science and Technology,Acadia University,Wolfville 4725,Nova Scotia,Canada;3.Department of Gynecology,Shenzhen Nanshan People’s Hospital,Shenzhen 518057,Guangdong,China;4.School of Pharmacy,Guangdong Medical University,Dongguan 523808,Guangdong,China)
关键词:
人工神经网络妇科疾病辨识数据挖掘中医证型
Keywords:
Artificial neural networksGynecological disease identificationData miningTCM syndromes
分类号:
R71
DOI:
10.3969/j.issn.1006-1959.2020.11.001
文献标志码:
A
摘要:
目的 建立基于人工神经网络编码数据挖掘技术的中医妇科病辨识数据分析方法。方法 检索中国期刊全文数据库(CNKI)、万方期刊数据库、维普中文期刊数据库,收集1980~2019年公开发表的关于妇科疾病多囊卵巢综合症(PCOS)的中医治疗与诊断方面文献,获取妇科辨证分型数据集,采用ANN模型对数据集进行量化分析。结果 通过模拟数据集ANN分析,建立了三层网络结构ANN模型,其中输入层包含15个输入神经元、隐含层包含4个神经元、输出层包含6个神经元,获得了多囊卵巢综合症中医妇科证素与证候之间的内在逻辑关系。重复训练及测试结果显示,中医证型的预测匹配率为100.00%。对15个输入协变量进行了参数重要性分析显示,精神状况>周期>身体上部>皮肤>舌象>经色>面色>大便8个症状的指标规范重要性大于50%。结论 基于人工神经网络技术建立了一种妇科疾病中医诊疗的证素-证型关系的神经网络方法,也为挖掘民族医学信息数据进行定性辨识、动态及多维数据的处理与分析提供一种有效途径。
Abstract:
Objective To establish a data analysis method for TCM gynecological disease identification based on artificial neural network coding data mining technology.Methods Retrieval of Chinese Journal Full-text Database (CNKI), Wanfang Journal Database, Weipu Chinese Journal Database, collection of publicly published literatures on TCM treatment and diagnosis of gynecological diseases polycystic ovary syndrome (PCOS) from 1980 to 2019, and access to gynecological syndromes Classification data set, ANN model is used to quantify the data set.Results Through ANN analysis of the simulation data set, a three-layer network structure ANN model was established, in which the input layer contains 15 input neurons, the hidden layer contains 4 neurons, and the output layer contains 6 neurons, and polycystic ovary syndrome is obtained The internal logical relationship between gynecological syndromes and syndromes in traditional Chinese medicine. Repeated training and test results show that the prediction matching rate of TCM syndromes is 100.00%. A parameter importance analysis of 15 input covariates shows that mental condition> period> upper body> skin> tongue> menstrual color> face color> stools are more than 50% of the index of 8 symptoms.Conclusion Based on artificial neural network technology, a neural network method of syndrome-syndrome relationship of traditional Chinese medicine diagnosis and treatment of gynecological diseases is established.

参考文献/References:

[1]袁金秋,刘雅莉,杨克虎.基于人工神经网络的数据挖掘技术在临床中应用进展[J].图书与情报,2010(3):95-98.[2]雷文平.试论人工神经络及其在机械工程领域中的应用[J].中国机械,2015(24):217-218.[3]邹慧琴,李硕,陶欧,等.RBF神经网络在中医药领域中的应用及其在SPSS 17.0软件中的实现[J].中华中医药学刊,2015,33(2):336-338.[4]Hassanzadeh H,Nguyen A,Karimi S,et al.Transferability of artificial neural networks for clinical document classification across hospitals:A case study on abnormality detection from radiology reports[J].J Biomed Inform,2018(85):68-79.[5]Egerer E,Siemonsen S,Erbguth F.Acute diseases of the brain and heart:A reciprocal culprit-victim relationship[J].Med Klin Intensivmed Notfmed,2018,113(6):456-463.[6] 许玉龙,盛梦园,王哲,等.几种数据挖掘方法用于中医证候分析的对比研究[J].中国中医药信息杂志,2019,26(12):97-102.[7]王玉蓉,曹云霞.精神、心理因素与多囊卵巢综合征[J].中国妇幼健康研究,2004,15(5):303-305. [8]蒋文星.数据挖掘在妇科PCOS中医证候及用药规律中的研究初探[D].广州中医药大学,2016.[9]黄文金,姚明龙,叶云金,等.关联规则在《中医妇科学》证素诊断中的应用[J].湖南中医杂志,2015(8):148-149.

更新日期/Last Update: 1900-01-01