Pre-eclampsia risk prediction method based on MLP multi-platform calibration

Abstract

The invention belongs to the technical field of computer-aided diagnosis, and relates to a preeclampsia risk prediction method based on MLP multi-platform calibration. The method comprises the following steps: collecting sample data to obtain sample features; carrying out conversion and blank filling on the basic features of the sample; carrying out normalization processing on placenta growth factors in the disease sample data and the disease-free sample data and the processed basic features; constructing an MLP model for multi-platform calibration by using an MLP, and PlGF values of different measurement platforms after normalization are calibrated to the same platform; constructing a preeclampsia risk prediction model based on a random forest by using basic features in the processed disease sample data and disease-free sample data and placenta growth factors; and performing preeclampsia risk prediction on a test sample by using the constructed prediction model. The problem that the existing preeclampsia prediction means cannot meet the requirement for early screening is solved, and a reliable auxiliary tool is provided for doctors to clinically diagnose preeclampsia cases.

Description

technical field [0001] The invention belongs to the technical field of computer-aided diagnosis, and relates to a preeclampsia risk prediction method based on multi-layer perceptron (MLP, Multilayer Perceptron) network multi-platform calibration. Background technique [0002] Preeclampsia is a kind of hypertensive disorder in pregnancy. It is a syndrome with complex and changeable symptoms that appears after 20 weeks of pregnancy and is very harmful to pregnant women and fetuses. It is clinically manifested as elevated blood pressure of pregnant women accompanied by a Or abnormalities of multiple organs / systems, such as proteinuria, abnormal renal function, and liver function damage. Preeclampsia is the second leading cause of death among pregnant women, with an annual incidence of about 8.5 million. The cause of its onset is still unclear, and it may involve various factors such as the mother, placenta and fetus. In all pregnancy periods, the prediction of the first trimes...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a preeclampsia risk prediction method based on MLP multi-platform calibration, to solve the problem that the current preeclampsia prediction methods cannot meet the needs of early screening, and to provide reliable clinical diagnosis of preeclampsia cases for doctors. auxiliary tool

Images