基于人工智能的不孕症中医辨证模型的构建与应用

Construction and application of TCM syndrome differentiation model of infertility based on artificial intelligence technology

目的:建立基于人工智能的不孕症中医辨证模型,为不孕症中医智能辨证模型的构建与应用提供方法和依据。方法:检索中国期刊全文数据库、万方期刊数据库、中医智库、古今病案云平台,收集关于不孕症的中医名医病案300例,建立不孕症病案中医信息数据库,采用经过超参数调优的随机森林(RF)、支持向量分类(SVC)、K-近邻(KNN)、人工神经网络(ANN)及统计学注意力神经网络模型(SANN)对数据集进行量化分析,建立不孕症辨证模型,采用五折交叉验证对模型进行评价,评价指标包括Accuracy、Precision、Recall、F1。结果:不孕症中医四诊信息为输入变量共86项,输出变量为不孕症中医证型共6项。5种模型的拟合效果较好,Accuracy、Precision、Recall、F1值均在0.77以上;其中SANN模型的准确率、查准率与查全率最高,Accuracy、Precision、Recall、F1分别为0.90、0.90、0.91、0.90,均高于其他算法模型,其参数的中医解释基本符合中医诊断原理。结论:基于SANN算法模型建立的不孕症中医辨证模型具有良好的诊断能力,将人工智能应用于不孕症中医辨证模型的构建与临床应用方法学可行,且具有较高的准确率。

Objective: To establish the TCM syndrome differentiation model of infertility based on artificial intelligence,so as to provide methods and basis for the construction and application of TCM intelligent syndrome differentiation model of infertility. Methods: The CNKI, Wanfang database, Chinese medicine think tank, Ancient and modern medical record cloud platform were searched. A total of 300 cases of medical records of famous TCM doctors about infertility were collected. And TCM information database of infertility medical records was established. Random forest(RF), support vector classification(SVC), K-nearest neighbor(KNN), artificial neural network(ANN) and statistical attention neural network(SANN) were used for quantitative analysis of the data set to establish the syndrome differentiation model of infertility. The model was evaluated by 5-fold cross validation. The evaluation indexes included Accuracy, Precision, Recall and F1. Results: There were 86 input variables and 6 TCM syndrome types as output variables. The Accuracy, Precision, Recall and F1 values of the five models were better than 0.77.The Accuracy, Precision, Recall and Recall of SANN model were the highest, and Accuracy, Precision, Recall and F1 were 0.90,0.90, 0.91 and 0.90 respectively, which were higher than other algorithm models, and the TCM interpretation of the parameters basically conformed to the principle of TCM diagnosis. Conclusion: The TCM syndrome differentiation model of infertility based on SANN algorithm model has good diagnostic ability, and the application of artificial intelligence in the construction and clinical application of TCM syndrome differentiation model of infertility is feasible and has high accuracy.