Acute disease occurrence probability assessment method based on ODE and GRUD

Abstract

The invention provides an acute disease occurrence probability assessment method based on ODE and GRUD and aims to solve the technical problem of low evaluation accuracy due to the fact that clinical index missing of time sequence data and non-uniform time intervals of sampling are not considered at the same time in the prior art. The method comprises the following implementation steps of: acquiring a time sequence data set, a timestamp sequence set and an acute disease tag set of multiple clinical indexes; obtaining an information attenuation characterization training sample set, an information attenuation characterization verification set sample and an information attenuation characterization test sample set; constructing a gating recursive unit network based on the ODE and the GRUD; performing iterative training on the gated recursive unit network based on the ODE and the GRUD; and obtaining an evaluation result of the occurrence probability of an acute disease. According to the method, the clinical index missing of the time sequence data and the non-uniform time intervals of the sampling can be solved, the influence of the clinical index missing is brought into a hidden state updating process, and the accuracy of the occurrence probability estimation of the acute disease is improved.

Description

technical field [0001] The invention belongs to the field of time-series signal processing, and relates to a method for evaluating the probability of occurrence of acute diseases, in particular to a method for evaluating the probability of occurrence of acute diseases based on ODE and GRUD, which can be used for multiple data sets with missing values ​​and uneven sampling time intervals. Analysis and processing of time series data of clinical indicators. Background technique [0002] In a clinical setting, data from various sensors and time-series data consisting of inspection and test results are of great significance for the assessment of the probability of a patient's acute illness. At present, medical staff generally evaluate the state represented by the data by focusing on several important clinical indicators and calculating the score, and judge whether there is an acute disease. Because the results of various sensors and inspections are complex and redundant to a cer...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to address the above-mentioned deficiencies in the prior art, and propose a method for evaluating the occurrence probability of acute disease based on ODE and GRUD, which is used to solve the lack of clinical indicators and sampling of time-series data in the prior art. The technical problem of low evaluation accuracy caused by uneven time intervals

Images