Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Disease risk prediction method and system based on multi-source graph neural network fusion

A disease risk, neural network technology used in the medical field

Pending Publication Date: 2022-07-22
BEIJING JIAOTONG UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a disease risk prediction method and system based on multi-source graph neural network fusion, which utilizes the multi-source disease relationship network to mine potential disease relationships, and fully learns and compares the historical diagnosis records of patients from the perspective of disease duration. Characterize and predict the risk of major chronic diseases to solve at least one technical problem in the above-mentioned background technology

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Disease risk prediction method and system based on multi-source graph neural network fusion
  • Disease risk prediction method and system based on multi-source graph neural network fusion
  • Disease risk prediction method and system based on multi-source graph neural network fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] The present embodiment 1 provides a disease risk prediction system based on multi-source graph neural network fusion, including:

[0034] The disease risk prediction set and the disease relationship network building module, by obtaining the patient's diagnostic record data, standardizing the patient's medical record, and integrating the diagnosis results of the patient's multiple visits from the perspective of the duration of illness to construct a disease prediction data set. Construct multi-source disease relationship network, including: disease co-occurrence network based on comorbidity, disease relationship network based on disease coding and disease network based on phenotype-genotype relationship.

[0035]The disease risk prediction module uses the constructed disease risk prediction model based on multi-source graph neural network fusion to predict the disease risk of patients and obtain the final disease risk; among them, the disease risk prediction model first u...

Embodiment 2

[0046] In this Example 2, a method for data integration and construction of a multi-source disease relationship network is proposed, and three disease relationship networks are constructed based on three disease relationship data, namely the disease co-occurrence network based on comorbidities, and the disease co-occurrence network based on ICD-10. Disease relationship networks and disease networks based on phenotype-genotype relationships, such as figure 2 shown.

[0047] (1) Disease network based on co-occurrence of comorbid diseases

[0048] A patient suffers from one disease and often suffers from multiple other diseases. This section constructs a disease co-occurrence network through the co-occurrence of the patient's disease. Each node of the network represents a disease, and the edge represents that the diseases corresponding to the nodes at both ends have a co-occurrence relationship, and the weight is the number of patients suffering from both diseases in the medica...

Embodiment 3

[0060] In this embodiment 3, a disease risk prediction model based on multi-source graph neural network fusion is proposed to achieve high-precision prediction of disease risk.

[0061] Disease risk prediction models based on disease duration data (such as image 3 (shown) contains three modules, A module is to extract the patient's historical disease characteristics; B module learns the node feature matrix of multiple disease networks with the help of a graph convolutional neural network model, and then calculates the high-dimensional characteristics of patients based on the disease network; Finally, the C module splices the patient characteristics learned by the A module and the B module, and uses the multi-layer perceptron to predict the patient's risk of new diseases in the future, as follows.

[0062] (1) Patient clinical characteristics learning module A. This module mainly includes ICD adjacency matrix activation and convolutional neural network (CNN) feature extractio...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a disease risk prediction method and system based on multi-source graph neural network fusion, and belongs to the technical field of medical treatment. In the method and the system, patient features are constructed based on disease duration information in combination with historical diagnosis information of a patient, and a disease risk prediction data set is formed; constructing a multi-source disease relation network, proposing disease network feature extraction based on a graph neural network, and performing patient disease feature matrix completion; a disease risk prediction model based on multi-source disease relation network fusion is provided, and high-precision prediction of disease risks is achieved.

Description

technical field [0001] The invention relates to the field of medical technology, in particular to a disease risk prediction method and system based on multi-source graph neural network fusion. Background technique [0002] With the popularization of electronic medical record system, disease prediction research has received extensive attention and made important progress in recent years. One of the goals of disease prediction research is to predict the risk value of a patient who may develop a certain disease in the future. It is mainly divided into two types of models: 1) Static prediction, that is, predicting a certain outcome without considering time constraints, similar to doctors 2) Dynamic disease prediction, that is, predicting a patient's condition by considering the patient's condition under multiple time nodes, such as using historical disease records to predict the future Risk of heart failure in half a year. [0003] Static disease prediction models are mainly b...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G16H50/30G06K9/62G06N3/04G06N3/08
CPCG16H50/30G06N3/08G06N3/045G06F18/241
Inventor 周雪忠田昊宇杨扩
Owner BEIJING JIAOTONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com