Prediction of abnormal pregnancy in pregnant women with Advanced maternal age and Pregestational Diabetes using Machine learning models

Abstract

Number of women with advanced maternal age is increasing along with some preexisting medical condition like diabetes. The objective of this paper is to use machine learning models to predict number of women above 35 with pregestational diabetes having high risk of abnormal pregnancy. Using data from Open access CTG database and Mother’s Significant Health dataset we compare women older than 35 with preexisting medical conditions to determine the risk prediction. The outcomes were compared between women 18 to 70 years and women above 35 years to see contrast in risk prediction and also the affected women already medical condition. Machine learning models of Random Forest, Logistic Regression and Decision tree were used to predict the outcomes. The target variable was set as pregestational diabetes. Random Forest and Logistic Regression performed equally better in both provided cases for women 18 to 70 years and women above 35 years of 87% and 83% respectively. Pregnant women above 35 having pregestational diabetes are at high risks of experiencing abnormal pregnancies. Future research and machine learning models may be able to detect the factors of the high risks involved for older pregnant women.