Development of Intelligent Application for Diagnosis of Gestational Diabetes Mellitus Disease in Pregnant Women

Abstract

Glucose intolerance during pregnancy or in women with reduced glucose tolerance after pregnancy termination is known as gestational diabetes mellitus. One of the main reasons pregnant women die is gestational diabetes mellitus (DMG). One of two factors can lead to diabetes: either insulin resistance (the body does not fully respond to insulin) or an autoimmune reaction (the body's defense system assaults the cells that make insulin). If there is an increase in both the 2-hour postprandial and fasting blood glucose levels between weeks 24 and 28 of pregnancy, DMG illness is manually diagnosed. Blood pressure and body mass index calculations are part of the physical examination in the early stages of pregnancy. Analyzing risk variables such age, hypertension, hyperlipidemia, prior history of gestational diabetes, family history of diabetes mellitus, and history of giving birth to infants weighing more than 4,000 grams (macrosomia) is essential during the anamnesis. Doctors also evaluate the results of the examination.  Developing a machine learning model to identify early gestational diabetes mellitus (DMG) in expectant mothers is the aim of this study. Artificial Neural Network (ANN) Backpropagation is the technique employed. Prior to being incorporated into the JST model, the dataset is cleaned and normalized. The optimal model for diagnosis is obtained by processing and testing the JST model. 94% accuracy is the resultant accuracy.