Abstract / Summary
Abstract Digital technologies and artificial intelligence (AI) have been increasingly incorporated into neonatal care, particularly in respiratory assessment. However, structured evidence regarding their clinical applications, performance, and limitations remains limited. This scoping review analyzed 35 studies published between 2019 and 2026 to map the use of digital technologies and AI in neonatal respiratory assessment. The findings showed that these approaches are mainly applied to predictive models, medical image analysis, and continuous physiological monitoring. Overall, the studies demonstrated potential for early diagnosis, prediction of respiratory outcomes, and clinical decision support. Nevertheless, the available evidence remains limited by methodological heterogeneity, lack of robust external validation, small sample sizes, and absence of protocol standardization. These findings suggest that, although promising, AI-based technologies are not yet ready for widespread clinical implementation, requiring prospective, multicenter, and standardized studies to validate their effectiveness and safety in neonatal care. Impact This review synthesizes the use of digital technologies and artificial intelligence in neonatal respiratory assessment, highlighting their expanding role in neonatal respiratory research and clinical decision support. AI-based predictive models demonstrate potential for early identification of critical respiratory outcomes, including bronchopulmonary dysplasia and the need for mechanical ventilation. Imaging-based approaches, such as chest radiography and lung ultrasound combined with deep learning, may improve diagnostic accuracy and reduce interobserver variability. Digital monitoring technologies enable continuous, noninvasive, and real-time assessment of neonatal respiratory parameters. The lack of external validation, methodological standardization, and evidence of clinical impact limits their application, reinforcing the need for multicenter studies and structured integration of these technologies into clinical workflows.