A Machine Learning Approach for the Prediction of Fetal Health using CTG
Astik Kumar Pradhan, Jitendra Kumar Rout, Aurobinda Bharat Maharana, Bunil Kumar Balabantaray, Niranjan Kumar Ray
Abstract
Cardiotocography or CTG is a technique for monitoring the fetal heart rate and uterine contractions during pregnancy. It is used to assess fetal well-being as well as to identify fetal distress early. The interpretation of a CTG can help determine whether a pregnancy is at high-risk or low-risk. An aberrant CTG may necessitate further study and, in some cases, intervention. These forecasts are evaluated in a real-time clinical judgment support framework which provides useful information that may be used to learn more about the fetal state. The current obstetric practice has made it possible to use various accurate and robust machine learning algorithms to classify fetal heart rate signals. Machine learning algorithms are becoming increasingly important in identifying ailments. The study aims to see how well machine learning models are suitable for predicting fetal health using CTG data. Several classifiers such as Logistic Regression, KNN, Random Forest, and Gradient Boosting Machine(GBM) have been used for the purpose, and their performance was evaluated in terms of accuracy, precision, recall, and F1-score. Experimental results show that Random Forest has the highest accuracy of 0.99.