Drug association graph construction method based on network science and convolutional neural network

Abstract

The invention discloses a drug association graph construction method based on network science and a convolutional neural network. The method includes the following steps that: feature extraction is performed on drug information, and a function association network between drugs is constructed through a convolutional neural network method on the basis of initial drug classifications given by pharmacists; a drug co-sale association network is constructed through a network science method on the basis of drug sales data; and the function association network and the drug co-sale association networkform a two-layer composite network, and comparison analysis is performed through the network science method, and a complete drug association graph is obtained; and the graph is visualized by means ofdimensionality reduction and graph embedding. With the drug association graph construction method based on the network science and the convolutional neural network of the invention, drug association relations can be automatically extracted from the drug sales data accurately and efficiently; and the association graph of drugs can be visually displayed, and users can know the drugs at a glance.

Description

technical field [0001] The invention relates to the field of medicine technology, in particular to a method for constructing a medicine association map based on network science and convolutional neural network. Background technique [0002] From the point of view of the object of use: it uses humans as the object of use to prevent, treat, and diagnose human diseases. Purposefully regulate human physiological functions, there are prescribed indications, usage and dosage requirements; from the perspective of usage: in addition to appearance, patients cannot recognize their inner quality, and many drugs need to be used under the guidance of doctors, not by patients' choice Decide. At the same time, various factors such as the method, quantity, and time of drug use largely determine the effect of its use. Misuse not only cannot "cure the disease", but may also "cause disease" and even endanger life safety. Therefore, medicine is a special commodity: type complexity: there are ...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the present invention provides a method for constructing a drug association map based on network science and convolutional neural network, which solves the problem that the existing network sales of drugs cannot accurately and efficiently extract drug associations from drug sales data automatically , and cannot visually display the problem of the association map between the main drugs

Images