Artificial Intelligence for Improved Maternal Healthcare: A Systematic Literature Review
Musa Chemisto, Tar J. L. Gutu, Kassim Kalinaki, Darlius Mwebesa Bosco, Percival Egau, Kirya Fred, Ivan Tim Oloya, Kisitu Rashid
Abstract
The integration of artificial intelligence (AI) in maternal health is a promising avenue for improving pregnancy, early childhood, and postnatal care. This systematic review analyzed 31 articles retrieved from Web of Science, PubMed, and Scopus, which were classified using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method and Mendeley referencing tool. Our interpretive study found that AI applications in maternal health can predict 48% of maternal complications using electronic medical records (EMR), 29% using medical images, 19% using genetic markers, and 4% using other medical features such as fetal heart rates and sensors. The accuracy of prematurity prediction using AI was 95.7%, while the XGBoost technique predicted neonatal mortality with 99.7% accuracy. The study underscores the potential benefits of AI in maternal healthcare and highlights the need for further research to improve maternal and child health outcomes, especially in resource-constrained sub-Saharan African regions where maternal mortality rates are significantly high.