Pipeline-Based Prediction of Gestational Diabetes Mellitus Using Synthetic Healthcare Data

Gestational Diabetes Mellitus (GDM) is a form of diabetes that arises during pregnancy in women with no prior history of the condition. It is caused by the body's inability to produce sufficient insulin to meet the increased metabolic demands of pregnancy. If left undetected or poorly managed, GDM can lead to serious health complications, including high maternal blood glucose, increased risk of cesarean delivery, neonatal hypoglycemia, and long-term susceptibility to obesity and type 2 diabetes for both mother and child. Early and accurate prediction of GDM is therefore essential for timely intervention and improved maternal-fetal outcomes. A pipeline-based predictive model is proposed, integrating Principal Component Analysis (PCA) for dimensionality reduction with a Random Forest classifier for robust prediction. The use of synthetic healthcare data enables the development and testing of the model in a controlled environment, enhancing reproducibility and scalability and its integration into clinical decision support systems for personalized prenatal care.