Medical record data structured analysis and processing method

Abstract

The invention discloses a medical record data structured analysis and processing method. The method comprises the following steps: the original medical record text is preprocessed so as to obtain thediagnosis result and the text without the diagnosis result; keyword recognition is performed on the text without the diagnosis result; the keyword feature vectors are extracted to perform machine learning and spatial visualization and classification analysis; and the correlation degree between different keywords and the diagnosis result is calculated according to the analysis results. According tothe method, the influence of the difference of the medical record format on the analysis can be reduced to the greatest extent, and the method has universal applicability and can be widely applied tothe analysis of various formats of medical records; classification analysis can be performed on the medical records of different departments under the guidance of the training data; and the basic information can be constructed for the subsequent medical record structured text and the reference for analyzing the correlation of the medical records can be provided.

Description

technical field [0001] The invention relates to the technical field of medical record data processing, in particular to a method for structured analysis and processing of medical record data. More specifically, it relates to a method for analyzing and processing medical record data structure based on natural language processing. Background technique [0002] Natural language processing (NLP) is one of the big applications of machine learning in recent years. Its research purpose is to explore how computers understand and use text and speech based on human natural language. The research goal of engaging in natural language processing is to design a machine algorithm similar to human beings through the analysis of the way humans understand language; the second is to use the more popular deep learning tools in recent years to directly analyze natural language texts . [0003] Medical record data contains rich text and numerical information. Now that the trend of electronic ...

AI Technical Summary

Problems solved by technology

Since natural language processing and medical treatment are not closely integrated at present, there are certain defects and deficiencies in the existing technology: First, although the medical records have been formatted as much as possible at the time of input, the format of medical records between different hospitals still exists The format of the old medical records in the past is even more varied, which is not conducive to the further development of medical research; second, there are few technologies for analyzing the relevant information sequence of Chinese medical records, which cannot provide convenience for medical researchers to explore the law of onset

Images