Electronic medical record named entity identification method, device and system, and remote terminal

Abstract

The invention provides an electronic medical record named entity identification method, device and system, and a remote terminal. The identification method comprises the following steps: S10, establishing a plurality of trained identification models; s20, acquiring an electronic case to be identified; s30, performing sequence labeling and data conversion on the electronic case in the step S20, and converting the electronic case into text data which can be recognized by a identification model; s40, respectively inputting the text data converted in the step S30 into a plurality of trained identification models, performing named entity identification, and respectively obtaining identification results of the plurality of models; and s50, performing fusion model operation on the plurality of identification results to obtain a final identification result after fusion. The method, device and system have the beneficial effects of improving the medical entity identification accuracy and further improving the working efficiency of checking the patient discharge information in a hospital, and is suitable for the field of natural language processing.

Description

technical field [0001] The invention relates to the technical field of natural language processing, in particular to a method, device, remote terminal and system for identifying named entities in electronic medical records. Background technique [0002] With the rapid development of information technology such as computers, the information construction of major hospitals is gradually improving, and a large amount of medical data is generated all the time. Although most medical records have been digitized, their role is still very important. If the data cannot be effectively used, the valuable value of the data will be greatly wasted. [0003] Electronic Medical Record (EMR) refers to the text, charts, images and other information generated by medical personnel through the medical information system during medical activities. It is a professional description of all aspects of the patient's body information and is a valuable medical record. [0004] When a patient is discharg...

AI Technical Summary

Problems solved by technology

[0006] However, the current named entity recognition methods for Chinese electronic medical records based on natural language processing usually include methods such as hidden Markov model (Hidden Markov Model, HMM), conditional random field (Conditional Random Field, CRF), and deep learning-based LSTM. Different methods have their own advantages and disadvantages; as the amount of data continues to increase and become more and more complex, existing methods can no longer meet the guarantee of high accuracy

Images