Multi-language medical term specification standardization system and method based on deep adversarial learning

Abstract

The invention discloses a multi-language medical term specification standardization system based on deep adversarial learning. The system comprises a standard medical term library, a file preprocessing module, a candidate term set generation module and a candidate term set rearrangement module. The file preprocessing module is used for unifying formats of medical term character strings in the real world and dividing long character strings into single Chinese and English tokens; the candidate term set generation module is used for screening first n candidate standard terms most similar to real-world medical terms from a standard medical term set; and the candidate term set rearrangement module is used for forming term pairs by using real-world medical terms and screened candidate terms, marking the term pairs into positive examples and negative examples according to the matching degree, taking the positive examples and the negative examples as input of a deep learning model, and generating an automatic term matching model through iterative optimization of the model.

Description

technical field [0001] The invention relates to a standardization system and method for multilingual medical terminology based on deep adversarial learning, and belongs to the field of medical technology. Background technique [0002] In the field of domestic medicine, especially in the writing of electronic medical records, there are many kinds of terms and the situation of non-standardization is serious. The previous scheme is to extract similar standard terms according to Chinese character strings after information extraction from electronic medical records, and then experts in the medical field Full manual or semi-automatic proofreading of these terms is performed. The standardization of medical terminology is time-consuming and laborious, and the efficiency is not high. [0003] After retrieval, Publication No. CN109446340A, a medical standard terminology ontology management system and method, equipment and storage medium, the patent discloses the term concept system, ...

AI Technical Summary

Problems solved by technology

[0007] 1. Due to the short history of Chinese medical informatization construction, the knowledge base and knowledge map resources containing Chinese medical standard terms, synonyms and other important semantic relationships, such as the relationship between diseases and drugs, are limited

[0008] 2. The term standardization method using synonyms as the main matching method is not enough to cover a large number of diverse term irregular expressions in real-world data

[0009] 3. The current work is mainly focused on the mapping between Chinese medical terminology expressions, lacking a flexible automatic model, which can flexibly support term mapping from Chinese to Chinese, Chinese to English, and Chinese to other languages, providing medical research, application, and medical information in my country. Obstacles to the internationalization of internationalized technology and products

Images