Triage method and system for nervous system disease

Abstract

The invention provides a triage method and system for a nervous system disease. The method provided by the invention comprises the steps: 1, carrying out the preprocessing of text data in a hospital electronic medical record, and extracting the symptoms and diagnostic data; 2, selecting the nervous system disease data for classification through a Bayes algorithm, and constructing a dedicated corpus; 3, constructing a training model based on a genetic algorithm, and carrying out the instructed learning of a training corpus; 4, modeling the classification after training, and constructing a triage knowledge graph of the nervous system disease.

Description

technical field [0001] The invention relates to the field of intelligent triage of nervous system diseases, in particular to a nervous system disease triage method and a nervous system disease triage system based on machine learning and cognitive computing. Background technique [0002] The concept of Computer Aided Diagnosis (CAD) was first proposed by Ledley in 1966. In 1972, Willcox attempted to apply Bayesian theory to computer identification of bacterial diseases. In 2007, Saeys reviewed the application of feature selection techniques in text mining in Bioinformatics published by Oxford University Press. He pointed out that although Feature Selection technology is widely used in the field of classification, its application in the field of biomedicine has just started. Medical text information has the characteristics of large dimensions and short sample length, and feature selection technology needs to be specific to medical data. Features are improved and optimized. ...

AI Technical Summary

Problems solved by technology

[0005] However, for the above-mentioned existing technologies, it is still not particularly effective in reducing the probability of calling the wrong number, reducing the probability of finding the wrong department / expert, etc.

Images