AI-augmented cardiopulmonary point-of-care ultrasound: a narrative review for pulmonary and critical care clinicians.

Point-of-care ultrasound (POCUS) is an increasingly utilised tool in pulmonary and critical care medicine, yet its effectiveness remains limited by operator dependency and the required cognitive demands of real-time image acquisition and interpretation. Artificial intelligence (AI) has emerged as a promising approach to augment POCUS by supporting image acquisition, interpretation and quantitative analysis. Given the rapid evolution of this field, clinicians must understand both the capabilities and limitations of AI-enabled POCUS. However, prior reviews have largely catalogued applications without focusing on how these tools integrate into bedside workflows. This narrative review examines AI in cardiopulmonary POCUS through a workflow-based lens relevant to pulmonologists and intensivists. We discuss the role of AI as a bedside coach, interpreter and quantifier, highlighting applications in lung and focused cardiac ultrasound, including pneumothorax detection, B-line analysis, left ventricular function assessment, and identification of other cardiopulmonary pathologies. Across domains, AI demonstrates strong performance in controlled settings, although variability persists in real-world environments due to differences in image quality, patient factors and dataset limitations. Future directions include interactive, clinician-in-the-loop systems, protocol-driven integration, and multimodal data synthesis. A clear understanding of these developments will be essential for clinicians as AI becomes increasingly embedded within bedside patient care.