Optimizing Prenatal Care with a Hybrid Model for Fetal Health

Fetal health classification involves assessing the wellbeing of a developing fetus. It is vital for early detection of risks, prevention of complications, optimizing prenatal care, reducing mortality and also for research and development for advancing medical knowledge. Early diagnosis and treatment of any disease process is essential for the health and safety of mother and baby. In this study, we created a hybrid model that effectively uses Cardiotocography (CTG) data to classify the fetus’s health, achieving remarkable accuracy rates of 98%. The remarkable accuracy rates were achieved through rigorous model training and validation, indicating the effectiveness of our techniques. The Hybrid model combines Random Forest, Support Vector Machine and Gradient boost powered by Artificial Neural Networks. Three skilled obstetricians divided the 2126 records in the dataset—21 attributes taken from cardiotocography exams—into three classes: normal, suspicious, and pathological. A hybrid model in fetal health classification offers improved accuracy, robustness, and feature representation by combining SVM, Gradient Boosting, Random Forest, and Artificial Neural Networks (ANN). It enhances generalization and adaptability, ensuring better prenatal care outcomes.