路志正教授治疗肺癌经验的数据挖掘及智能诊疗辅助研究

Research on data mining and intelligent assistant model of diagnosis and treatment of Lu Zhizheng′s experience in treating lung cancer

目的探索路志正教授治疗肺癌的用药规律。方法采集2011年1月1日至2019年12月31日路志正教授接诊的有效患者154例,合计642诊次,涉及8大虚证和8大实证,共22种证型组合和642个处方中的1511味中药,建立数据库,采用Python 3.8对证型和中药进行频次、关联规则和复杂网络分析。结果证型频次出现比例最高者为肺脾气虚、痰瘀互结证,利用关联规则分析,验证已知并发现未知中药配伍。通过复杂网络研究,构建路氏经验方治疗肺癌经验的智能诊疗辅助模型,初步实现了辨证明确的前提下,自动生成参考基本方、参考配伍。结论主要证型数据挖掘结果基本符合路氏经验制方原则,存在少许差异,非主要证型的基本方和配伍规律有待进一步挖掘,研究尚未发现肿瘤细胞分型与中医证型之间的对应关系,此恐与样本量有限相关。

Objective To explore the medication rule of Professor Lu Zhizheng in the treatment of lung cancer.Methods 154 cases of effective patients treated by Professor Lu Zhizheng from January 1,2011 to December 31,2019 were collected,with a total of 642 times,involving 8 dificency syndromes and 8 excess syndromes,a total of 22 syndrome type combinations and 1511 herbs in 642 prescriptions.The database was established,and the frequency,association rules and complex network of syndrome types and traditional Chinese medicine were analyzed by Python 3.8.Results The highest proportion of syndrome types was lung spleen qi deficiency,phlegm and blood stasis syndrome.The known and unknown compatibility of traditional Chinese medicine was verified by association rule analysis.Through the study of complex network,the intelligent diagnosis and treatment assistant model of Lu′s empirical prescription for lung cancer was constructed,and the basic formula and compatibility for reference were automatically generated under the premise of clear syndrome differentiation.Conclusion The results of data mining of main syndrome types basically conform to the principle of Lu′s empirical formula,and there are some differences.The basic prescriptions and compatibility rules of non main syndrome types need to be further explored.The corresponding relationship between tumor cell classification and TCM syndrome types has not been found,which is related to the limited sample size.With the further expansion of sample size in the future,it is possible to find new rules.