Abstract:
Objective: To explore the applications of artificial intelligence (AI) in neonatal intensive care units (NICUs), identifying key trends in AI-driven technologies and their roles in diagnosis, monitoring, and treatment of neonatal conditions.
Methods: A PRISMA-guided systematic review was conducted across MEDLINE, EMBASE, Cochrane, and IEEE Xplore, covering studies published between January 2013 and December 2023. A total of 318 studies were initially retrieved. After removing 61 duplicates and screening 257 articles by eligibility criteria, 64 studies were assessed for full-text eligibility, leading to the final inclusion of 41 studies.
Results: The predominant domains for AI application were cardiovascular conditions (n=9, 21.9%), followed by neural/brain conditions (n=8, 19.5%), respiratory difficulties (n=8, 19.5%), infections (n=6, 14.6%), digestive functions (n=2, 4.9%), and microvascular diseases (n=1, 2.4%). Additionally, 6 studies focused on monitoring systems or body positioning (categorized as "Not Disease"), and 1 study (2.4%) addressed mortality prediction. Regarding AI application purposes, prognosis (n=23, 56.1%) was the most common, followed by classification (n=14, 34.1%), monitoring (n=5, 12.2%), and symptom forecasting (n=1, 2.4%).
More than 70% of studies (n=29, 70.7%) lacked a validation procedure, highlighting a critical gap in methodological rigor.
Conclusions: Our findings underscore the significant benefits of AI in NICUs, including improved patient outcomes and enhanced operational efficiency. However, challenges such as data privacy, algorithm interpretability, and ethical considerations must be addressed for responsible AI deployment in neonatal care. We highlight future directions, emphasizing interdisciplinary collaboration, adherence to reporting guidelines, and further research to enhance the reproducibility and clinical integration of AI in NICUs. This study reinforces the transformative potential of AI in shaping the future of neonatal healthcare
