Disease prediction method based on heterogeneous medical knowledge graph and related equipment

Abstract

The embodiment of the invention relates to the technical field of medical treatment, and discloses a disease prediction method based on a heterogeneous medical knowledge graph and related equipment; and a processor of the equipment is used for executing the steps: obtaining a heterogeneous graph of a heterogeneous medical knowledge graph of historical diagnosis and treatment data of a user, inputting the heterogeneous graph into a graph attention neural network model to obtain a first node representation of each node in the heterogeneous graph; inputting a first user representation determinedaccording to the first node representation and the historical diagnosis and treatment data into a first multi-layer perceptron model to obtain a code representation, and performing pre-training according to the code representation to obtain a first disease prediction model; inputting the historical diagnosis and treatment data into the first disease prediction model to obtain a second user representation, and performing training according to outcome data obtained by inputting the second user representation into a second multilayer perceptron model to obtain a second disease prediction model; and inputting the target medical data into the second disease prediction model to obtain predicted outcome data. The invention relates to a blockchain technology. The data can be stored in a blockchain.

Description

technical field [0001] The present invention relates to the field of medical technology, in particular to a disease prediction method based on a heterogeneous medical knowledge map and related equipment. Background technique [0002] Disease prediction is to infer the risk of future diseases or clinical events based on the user's historical information. In order to solve the problem that data-driven disease prediction methods are easily affected by limited data volume and large data deviation, many works try to integrate medical knowledge maps into the disease prediction process. , to improve prediction accuracy and consistency with existing medical knowledge. [0003] However, at present, it is basically based on simple isomorphic knowledge graphs. The existing simple isomorphic knowledge graphs are difficult to use to capture complex heterogeneous information in heterogeneous medical knowledge graphs, and it is impossible to distinguish the coexistence of heterogeneous kno...

AI Technical Summary

Problems solved by technology

[0003] However, at present, it is basically based on simple isomorphic knowledge graphs. The existing simple isomorphic knowledge graphs are difficult to capture complex heterogeneous information in heterogeneous medical knowledge graphs, and cannot distinguish the coexistence of heterogeneous knowledge knowledge graphs. Positive and negative relationships, e.g. diagnosis and drug have both indication and contraindication relationships

The heterogeneous medical knowledge graph is huge, and the end-to-end training system based on limited data is prone to underfitting, which may lead to inaccurate prediction results

Images