Gestational diabetes mellitus screening method based on machine learning and physical examination data

Abstract

The invention provides a gestational diabetes mellitus screening method based on machine learning and physical examination data, and the method is applied in the technical field of disease screening.The method includes the following steps that first physical examination data is acquired and pretreated to obtain test samples; a target model obtained by the gestational diabetes mellitus screening method based on machine learning and physical examination data serves as a training mode for training the obtained test samples and outputting a predict result, when the predict result shows 1, it means that a pregnant suffers from gestational diabetes mellitus, and when the predict result is 0, it means that the pregnant does not suffer from gestational diabetes mellitus. According to the embodiment, a gestational diabetes mellitus screening model suitable for practical application scenarios is constructed via the LightGBM algorithm based technology and the characteristics of health data. Themodel can predict whether the pregnant suffers from gestational diabetes mellitus according to the genetic data and physical examination indexes of the pregnant, thus assisting doctors in diagnosis and treatment, and provides solutions for intelligent medical treatment. The purpose of improving the utilization rate of public medical resources is achieved.

Description

technical field [0001] The invention relates to the technical field of gestational diabetes screening, in particular to a screening method for gestational diabetes based on machine learning and physical examination data. Background technique [0002] There are two cases of diabetes during pregnancy, one is diabetes diagnosed before pregnancy, which is called "diabetes combined with pregnancy"; the other is diabetes that occurs or is diagnosed during pregnancy with normal glucose metabolism or potential impaired glucose tolerance before pregnancy , also known as "gestational diabetes mellitus (GDM)". More than 80% of pregnant women with diabetes have GDM, and the incidence of GDM in my country has reached 17.5% to 18.9%. With the development of society and economy, there are more and more high-risk pregnant women such as advanced age, pre-pregnancy overweight and obesity, and the incidence of GDM in my country has increased significantly in recent years. Most GDM patients c...

AI Technical Summary

Problems solved by technology

The prediction accuracy of the currently used methods is not high or it is impossible to achieve high accuracy and high operating speed at the same time

Images