Data selection method for diagnostic predictions

Abstract

The invention relates to the technical field of data message distribution, in particular to a data selection method for diagnosis prediction. The method comprises the following steps: S1, analyzing an electronic medical record and obtaining text data; S2, extracting the medical record data and diagnosis data; S3, uploading the data to a medical record library, and searching for related data; S4, comparing the medical record data and the diagnosis data; S5, performing data similarity arrangement; S6, obtaining multiple sets of related diagnosis schemes, wherein the text data refer to complete diagnosis medical record data information of a patient in a special department. According to the method, the tedious steps of medical record modeling and machine learning are avoided, the implementation cost of a diagnosis prediction scheme is reduced, multiple diagnosis schemes are provided when doctors diagnose special medical records or medical records containing complications, the situation that the complex medical records are missed or mistakenly diagnosed is avoided, and meanwhile, the searched comparison medical records provide treatment methods and treatment effects of previous medical records for doctors, so that the doctors can conveniently treat patients after diagnosis and prediction.

Description

technical field [0001] The invention relates to the technical field of data selection for diagnostic prediction, in particular to a data selection method for diagnostic prediction. Background technique [0002] In the medical field, medical record diagnosis is the standard definition for determining a patient's illness, and is one of the important materials for hospital management, medical treatment, scientific research and teaching, and has important scientific value, especially with the reform and standardized management of the medical system, based on The classification of diseases diagnosed by medical records has become an important basis for cost accounting and payment. [0003] The Chinese patent application number is CN 108597614 A, which discloses a decision-making method for auxiliary diagnosis based on Chinese electronic medical records, including: first cleaning the disease and symptom information collected in Chinese electronic medical records, and then mining th...

AI Technical Summary

Problems solved by technology

However, in the diagnosis process, in order to ensure the accuracy of diagnosis, a large number of case models need to be imported into the database, and a large number of data models must be imported into the intelligent machine library for learning and programming before they can be used, which greatly improves the accuracy of diagnosis. Forecast costs and maintenance costs

Images