Artificial Intelligence non-invasive methods for neonatal jaundice detection A review (2025) Fati Oiza Salami

Description

This review analyzes AI-driven non-invasive methods (Machine Learning and Deep Learning) for detecting neonatal jaundice, a common and potentially fatal condition. Traditional diagnosis (Total Serum Bilirubin testing) is often invasive and delayed. The review finds that AI models evaluating complex patterns in neonatal skin color achieve over 90% accuracy. Mobile-based applications using smartphone cameras also demonstrate satisfactory outcomes in field settings, providing a practical alternative for resource-constrained areas. The paper evaluates the impact of AI-based strategies in revolutionizing neonatal care and suggests future research areas.