Drug group mining method and device of target disease, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a drug group mining method and device of a target disease, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence. The drug group mining method for the target disease comprises the following steps: acquiring target medical data of the target disease; performing vector conversion processing on the drug group data to obtain a drug group vector; obtaining a target sentence vector according to the prescription data and the drug group vector; obtaining hospitalization duration frequency classification data according to the medication frequency data; obtaining clustering parameter information according to the hospitalization duration frequency classification data, and performing clustering processing on the target sentence vector to obtain clustering result data; carrying out survival analysis processing according to the clustering result data to obtain survival curve information; and performing drug mining processing from the drug group data according to the survival curve information to obtain a target drug group mining result for the target disease, and through the technical scheme provided by the embodiment of the invention, the effectiveness of the drug mining result can be improved.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, and in particular, to a method and device for mining a drug group of a target disease, an electronic device, and a storage medium. Background technique [0002] The current drug mining methods for traumatic brain injury and other diseases mainly include the following two categories: (1) data mining methods based on data models, (2) survival analysis methods based on survival time and survival status. [0003] Among them, the current drug mining method (1) is still in the stage of drug statistics, which is greatly affected by the prescriptions included in the literature. The richness of prescriptions determines the integrity of the database. For example, the drug mining method with clinical symptoms as the main research body, only Considering the corresponding relationship between symptoms and drugs and the frequency of use of drugs corresponding to symptoms, the factors co...

AI Technical Summary

Problems solved by technology

[0003] Among them, the current drug mining method (1) still stays in the drug statistics stage, which is greatly affected by the prescriptions included in the literature, and the richness of prescriptions determines the integrity of the database. Considering the corresponding relationship between symptoms and drugs and the frequency of use of drugs corresponding to symptoms, the factors considered are relatively single, resulting in poor effectiveness of drug mining results; and the data model has defects that are too strict, such as requiring sample data to obey Multivariate normal distribution, the same covariance matrix, etc., which lead to limited computing efficiency of mining algorithms; the current drug mining method (2) only considers the relationship between survival time and survival status, and then uses the survival curve to express the survival probability, resulting in Poor validity of drug mining results

[0004] In summary, the current drug mining methods have the problems of limited computational efficiency and poor effectiveness of drug mining results.

Images