Artificial intelligence-assisted airway evaluation and imaging-based decision-making in head and neck anesthesia: a narrative review

Background and Objective: Artificial intelligence (AI) has become an integral part of contemporary anaesthesiology, particularly in complex clinical scenarios in the head and neck. With oral and maxillofacial anaesthesia and head and neck and otorhinolaryngologic surgery, practitioners frequently encounter difficulties with airway management, anatomical variations, tumour distortion, limited mouth opening, and procedure-specific risks. Imaging modalities including ultrasound and computed tomography (CT) in conjunction with clinical information are essential for pre-anesthetic assessment and perioperative planning. This narrative review provides a clinically orientated overview of the current use of AI in head and neck anaesthesia, focusing on airway assessment, imaging-based risk stratification, ultrasound-guided regional anaesthesia, and imaging-assisted decision support.

Methods: A literature search was performed using PubMed, Google Scholar, Web of Science, and Scopus, including studies published from January 1997 to June 2025. Search terms included “artificial intelligence”, “machine learning”, “deep learning”, “head and neck anesthesia”, “airway management”, “computed tomography”, “ultrasound”, and “ultrasound-guided anesthesia”. Original studies and review articles published in English and addressing AI applications in airway assessment, imaging analysis, ultrasound-guided procedures, and perioperative decision-making were included.

Key Content and Findings: Current evidence suggests that machine and deep learning models may enhance prediction of difficult airway by incorporating clinical and imaging data, particularly anatomical features derived from CT. AI-aided CT analysis has shown promising application for objective airway assessment and risk stratification, and ultrasound-based applications have been explored for anatomical landmark identification, image interpretation, and needle guidance. While these technologies have great technical potential, the majority of the available studies remain exploratory and are largely based on retrospective datasets, proof-of-concept systems, or technical feasibility studies. External validation is still limited, especially in patient cohorts typically seen in oral and maxillofacial anaesthesia.

Conclusions: AI can improve airway assessment, imaging interpretation and perioperative decision-making in head and neck anaesthesia. While many applications are still in the early stages of clinical implementation, current evidence suggests that AI may become an increasingly valuable adjunct in the management of anatomically complex patients. Future research should focus on prospective validation and clinically relevant applications, especially in the field of oral and maxillofacial anaesthesia.