Foetal health monitoring is a vital component of prenatal care, ensuring the well-being of both mother and child. Traditional techniques such as cardiotocography (CTG) are widely used, however, they are often limited by subjective interpretation and time inefficiencies. There is also limited literature that presents guidelines on the application of machine learning in Foetal health status prediction, particularly addressing challenges such as data variability, model generalizability, and algorithmic bias. This paper presents findings from a Systematic Literature Review (SLR) that followed the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) model. The SLR examined 42 peer-reviewed studies published in between 2021 and 2024 from Google Scholar and IEEE Xplore databases. The findings indicate that while machine learning models like random forests and neural networks achieve high predictive accuracy, their performance is often limited by data variability, lack of generalizability across diverse populations, and algorithmic bias. By addressing these challenges, Machine Learning can improve foetal health outcomes, ensuring timely and accurate interventions. These findings inform healthcare policy-makers on how to intervene and manage foetal health hence reducing the mortality and morbidity of mother and/or child. Researchers interested in martenal and foetal health will also benefit from the findings of this paper through adhering to the guidelines provided in the resultant framework to improve the management of the health of the foetus.

Keywords

Foetal Health Status, Prediction, Machine Learning, Fetus Health Monitoring, Foetal Well-being