Structured method for disease data and thyroid cancer risk prediction system

Abstract

The invention discloses a structured method for disease data and a thyroid cancer risk prediction system. The method comprises the following steps: acquiring a disease big data queue from a distributed medical information database according to a medical dictionary, wherein the disease big data queue comprises identity information, disease information and a data source; extracting a sample data setbased on the disease big data queue, and subjecting the sample data set to structuralization according to a disease classification standard; and based on the structured sample data set, subjecting the remaining data in the disease big data queue to structuralization. Besides, based on the structured disease big data queue, a thyroid cancer data queue is extracted, and an onset risk prediction model is established. According to the invention, structuralization of massive medical big data is realized, and prediction of thyroid cancer attack probability can be realized.

Description

technical field [0001] The invention belongs to the technical field of medical big data processing, and in particular relates to a disease data structuring method and a thyroid cancer risk prediction system. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] The incidence of thyroid cancer has increased significantly worldwide over the past few decades, and following this trend, thyroid cancer will become the fourth most common cancer in the United States by 2030. The incidence of thyroid cancer in my country is also showing a rapid growth trend, and it has become a malignant tumor that seriously affects the health of the population, especially the health of women. Although the incidence of thyroid cancer has been rising steadily, the mortality rate has gradually stabilized, and even the mortality rate of thyroid cancer has declined slightl...

AI Technical Summary

Problems solved by technology

Combined with the characteristics of medical data and related research situations, there are still some problems to be studied in the current work, which are reflected in two aspects: 1) Using homogeneous information network graphs to construct medical data description models ignores the characteristics of data heterogeneity, resulting in a lot of problems. A large degree of semantic loss; 2) The existing reasoning models are mainly quantitative reasoning, lack of research and application of qualitative reasoning based on causality

[0005] Insufficiency of the medical data description model: Among the many characteristics, the heterogeneity of the medical data is the most prominent, the coverage of the data is unprecedented, the format of the data is varied, and the sources of the data are also complicated

However, in the traditional description method based on homogeneous graphs, many different types of data are modeled as indistinguishable nodes, and the connections between nodes are not distinguished. This kind of processing often ignores the differences between data types, and even leads to many large loss of semantics

[0006] Lacking a comprehensive judgment based on causality, the traditional forecasting model only gives the final risk probability, and lacks the analysis of the influence degree of influencing indicators, making it impossible for users to know which indicators of their own led to this result

Images