Study research on multi-level representation of syndrome elements in TCM intelligent auxiliary diagnosis system
10.3760/cma.j.cn115398-20200603-00043
- VernacularTitle:面向中医智能辅助诊断系统的证候要素多层次表示方法研究
- Author:
Yingjie SHI
1
;
Wen HUANG
;
Pan ZHAO
;
Chengyuan LIU
;
Zongyou LI
Author Information
1. 中国中医科学院中医药信息研究所2018级博士研究生，北京 100700
- From:
International Journal of Traditional Chinese Medicine
2020;42(10):1001-1007
- CountryChina
- Language:Chinese
-
Abstract:
Objective:This paper studied the classification of syndrome and the multi-level expression method of syndrome elements, to realize the processing of syndrome name wiht computer, which could be applied to the research on traditional Chinese medicine (TCM) intelligent auxiliary diagnosis system.Methods:With the guidance of TCM theory combining the syndrome differentiation and syndrome elements theory, the syndrome elements are divided into a more fine-grained way, and the hierarchical division method and multi-level expression method of syndrome elements are put forward. Than to conduct artificial analysis on the syndrome of gynecological diseases to verify the rationality of multi-level expression of syndrome elements. Based on the multi-level expression method of syndrome elements and pattern extraction method of pattern matching,form the extraction method of syndrome elements and through the batch processing of the Clinical Diagnosis and Treatment Term of TCM& Syndrome Part to verify the accuracy of the batch processing methodexpression. Results:Guided by TCM theory, this study divided the constituent elements of syndrome names into two levels and five types containing disease location elements, disease nature elements which includes basic material, etiology, pathological state and conjunctions, based on which the syndrome splitting process is constructed. Based on this, the process of syndrome name elements separation is constructed. Artificial analysis of TCM syndrome names of menstrual diseases and infertility shows that all syndrome names can be divided into one or more disease location elements, disease nature elements, and conjunctions based on that splitting process. The accuracy rate of computer batch processing is as high as 71.4% based on that splitting process.Conclusions:The multi-level representation method of syndrome elements provides a theoretical framework for mass data processing by computer, improves the efficiency and accuracy of data processing, provides a theoretical basis for the construction of training data set of syndrome differentiation model, provides support for TCM syndrome differentiation reasoning model, and can be applied to the research of TCM intelligent auxiliary diagnosis system.
