Electronic medical record phenotype extraction and phenotype name normalization method and system

Abstract

The invention discloses an electronic medical record phenotype extraction and phenotype name normalization method. The method comprises the following steps: phenotypic extraction: taking natural statements of a medical record text as original data, and adopting Bi-; Named entity recognition is carried out on the LSTM model and the CRF model, and a phenotypic entity class is extracted; And phenotype standardization is carried out, an LSTM encoder is adopted to encode each phenotype, cosine similarity between non-standard phenotype codes and standard phenotype codes in the medical records is calculated, and the non-standard phenotype codes are mapped to the phenotype with the highest cosine similarity. The invention further discloses an electronic medical record phenotype extraction and phenotype name normalization system. According to the method, the named entity identification accuracy, the recall accuracy and the phenotypic mapping accuracy in the electronic medical record are improved; The labor consumption in the medical record structuring process is avoided, and the medical record structuring efficiency is improved; The method can more efficiently and accurately serve medical data mining, clinical decision support, clinical risk assessment and the like.

Description

technical field [0001] The present invention relates to the technical field of medical text data processing, in particular to a method and system for extracting phenotypes from electronic medical records and standardizing phenotype names based on deep learning. Background technique [0002] In conventional medical electronic records, the main carrier of information is natural language, such as image reports, medication records, disease course reports, and medical record inspection reports. It can be said that these natural language texts contain the main clinical information in the process of patient diagnosis and treatment. On the one hand, in recent years, hospitals at all levels in my country have gradually adopted electronic medical record management systems to replace traditional handwritten medical records, thus accumulating more and more medical electronic records. On the other hand, with the deep integration of information technology represented by big data and arti...

AI Technical Summary

Problems solved by technology

Considering the huge amount of data, if relying on manual structuring, the economic cost will undoubtedly be very high

Images