Traditional Chinese medicine electronic medical record structuring method and terminal

Abstract

The invention provides a traditional Chinese medicine electronic medical record structuralization method and terminal, and the method comprises the steps: inputting traditional Chinese medicine electronic medical record text data into a trained BertCRF model, and obtaining a traditional Chinese medicine entity; performing boundary correction on the traditional Chinese medicine entities based on atraditional Chinese medicine entity library to obtain corrected entities; decomposing symptom entities in the corrected entities based on a symptom decomposition element dictionary tree and a maximumforward matching algorithm to obtain decomposed entity elements; and querying in an element map based on the entity elements to obtain standard element nodes corresponding to the entity elements, andobtaining a standard symptom group through the standard element nodes. The boundary problem of part of traditional Chinese medicine entity identification is solved, and the accuracy and recall rate ofentity identification are improved; according to the scheme, a traditional Chinese medicine symptom normalization method is adopted, and symptom normalization can be accurately carried out.

Description

technical field [0001] The invention relates to the technical field of data structuring, in particular to a method and terminal for structuring electronic medical records of traditional Chinese medicine. Background technique [0002] Due to the complexity and particularity of the text representation of TCM medical records, there are many proper nouns and representations in TCM medical records, which are often quite different from common Chinese representations. Names, especially for the standardization of names for symptoms. [0003] In the existing schemes, models such as HMM and CRF are mainly used for entity recognition of electronic medical records. The method of HMM, CRF and other models for entity recognition of electronic medical records is a commonly used method in western medicine electronic medical record entity recognition, but it cannot achieve the expected effect in the field of traditional Chinese medicine, because the training corpus does not contain a large ...

AI Technical Summary

Problems solved by technology

However, the minimum edit distance and the Jaccard similarity coefficient are all based on string similarity calculation methods, which cannot handle the situation where the characters have completely different meanings but the same meaning, such as "anorexia" and "loss of appetite". high approximation

The method based on the word vector, because the word vector comes from the word vector obtained by the language model based on the context training, will cause the symptom word vectors that often appear together to have a high degree of similarity, which will lead to similarity calculation errors

Images