Comparitive Analysis of Machine Learning Models for Prediction of Fetal Health

The World Health Organization estimates that about 6.3 million prenatal deaths occur annually across the globe. However, with timely and accurate diagnosis of certain complications such situations can be avoided or can be resolved with proper medical attention. Cardiotocogram (CTG) is one of the tools used for recording fetal heart rate and uterine contractions in the womb which can provide insights into the fetus' health. The process of interpreting CTGs is time consuming and requires domain expertise. This process can be simplified by using machine learning algorithms which can not only predict whether the fetus is healthy, but also provide insights to factors influencing the health of the fetus. This paper aims to analysis different machine learning algorithms for interpretation of CTGs to classify fetal health as normal, suspected, or pathological.