Hierarchical prediction method and system for electronic medical records and computer equipment

Abstract

The invention discloses a hierarchical prediction method and system for electronic medical records and computer equipment. The method comprises the following steps: 1) constructing an electronic medical record data set of a trauma patient; 2) performing category label division, selecting a part of data as a training set, and taking the remaining data as a label-free corpus after labels are hidden; 3) further training the pre-trained language model by using the untagged corpus to obtain a further trained language model; 4) using the language model obtained by further training to represent a training set sample, and performing classification training to obtain a final classification prediction model; and 5) inputting to-be-graded electronic medical records into the grading prediction model to obtain a grading result of the electronic medical records. According to the method, the electronic medical record data can be utilized to perform accurate grading prediction on the triage grade of the first-aid trauma patient, so that information with reference value is provided for further processing of doctors, the grading prediction accuracy can be improved, and the workload of medical personnel is reduced.

Description

technical field [0001] The invention relates to the technical field of smart medical care and information processing, in particular to a method, system and computer equipment for graded prediction of electronic medical records. Background technique [0002] In the existing emergency services, usually the operator or triage staff will conduct initial inquiries based on the patient's condition and record the patient's electronic medical record. These medical records contain the patient's personal information, current medical history, past history and other information. The evacuation hospital often judges the triage level of the patient based on this part of the electronic medical record by the doctor. The triage level is generally divided into 4 levels according to the severity, endangered, critical, emergency, and non-emergency. The emergency department will use the severity level as a reference data for further processing. Safe and effective pre-screening triage can ensu...

AI Technical Summary

Problems solved by technology

[0003] Manual accurate triage level judgment by medical personnel has the disadvantages of heavy workload and deep dependence on doctors' professionalism and experience; using electronic medical record data and classification models to accurately grade and predict the triage level of emergency trauma patients, It can also provide a valuable reference for further treatment by doctors, and it can reduce the workload of medical staff, which is of great significance; but there is no public and reliable solution.

Images