Scoping review: AI and digital technology in neonatal respiratory assessment

AI-based predictive models and imaging (chest X-ray, lung ultrasound) show potential for early diagnosis and outcome prediction in neonatal respiratory care, but lack external validation, standardization, and multicenter evidence.

• Analysis of 35 studies (2019–2026) focused on predictive models, medical image analysis, and continuous physiological monitoring.
• AI applications included early identification of bronchopulmonary dysplasia and need for ventilation, deep learning on chest radiographs and lung ultrasound, and digital monitoring tools.
• Methodologic heterogeneity, small samples, and absence of robust external validation limit clinical readiness; prospective multicenter studies are needed.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.