Fetal health classification with predictive algorithm by using Ensemble Model

Department of CSE, Sreenidhi Institute of Science and Technology, India

Abstract

Fetal health assessment is essential for ensuring the well-being of both the mother and fetus during pregnancy. Cardiotocography (CTG) is a widely used technique that monitors fetal heart rate (FHR) patterns and uterine contractions, providing critical insights into fetal health. However, interpreting CTG data is often subjective and error-prone. This study suggests a new way to classify fetal health using machine learning and a CTG dataset anyone can access. The dataset has important features that doctors look at to check fetal health. These include long drops in heart rate unusual short-term changes how often the long-term pattern is off how spread out the data is, the middle value, the average long-term change, the most common value, and times when the heart rate speeds up. The study uses two strong machine learning methods XGBoost and LightGBM, to sort the data. It also tries combining these methods to get better results. Tests show that mixing the two methods works best getting over 93% accuracy in telling how healthy a fetus is. This shows that using XGBoost and LightGBM together makes good use of the chosen features leading to more accurate sorting. By bringing machine learning into fetal health checks, this study gives doctors a trustworthy tool to help them make quick accurate diagnoses. This can lead to better outcomes for pregnancies and lower the chance of problems