O uso da Inteligência Artificial como ferramenta de decisão clínica em Neonatologia, com ênfase na otimização da Farmacoterapia: uma revisão integrativa

Abstract

The complexity of care in Neonatal Intensive Care Units demands greater precision and speed in decision-making. The objective of this work was to conduct an integrative literature review on the use of Artificial Intelligence (AI) as a clinical decision-making tool in neonatology, with emphasis on optimizing pharmacotherapy, early diagnosis, and outcome prediction. The search was conducted in the PubMed and LILACS databases, resulting in the analysis of 53 studies. The results demonstrated that Machine Learning and Deep Learning algorithms, such as XGBoost and Convolutional Neural Networks, achieved high accuracy in adjusting antimicrobial doses, screening for Retinopathy of Prematurity, and detecting late-onset sepsis. They are also important in the detection and management of seizures, as well as in the therapeutic optimization with surfactants, caffeine, and probiotics. In conclusion, AI has established itself as a transformative and auxiliary resource, capable of processing large volumes of data to offer precision medicine, ensuring greater safety and therapeutic effectiveness, although it must always operate under the supervision of human clinical judgment.