Abstract Preoperative airway assessment relies heavily on bedside examination, yet traditional tests are often poorly standardised and limited in predicting difficult facemask ventilation, laryngoscopy or intubation. The data from the United States closed claims and the 4 th National Audit Project indicate that although predictors of difficult intubation are frequently present, failure to anticipate airway challenges contributes to inappropriate management in most cases. Conventional assessments are particularly limited in evaluating the pharyngo-laryngo-tracheal region, an anatomical ‘blind spot’ where tumours, infections, inflammation or trauma may compromise airway patency. Imaging modalities such as radiographs, computed tomography and magnetic resonance imaging can reveal anatomical distortions, lesion extent and secondary tissue involvement that may indicate difficulty with intubation. Dynamic airway mapping techniques, including flexible nasal endoscopy, virtual endoscopy and Point-of-Care UltraSound, offer intraluminal and functional insights to guide safer, individualised airway management. The emerging use of artificial intelligence (AI) in integrating imaging data to develop predictive models may further enhance preoperative airway assessment to guide safe airway planning. This narrative review discusses the current use and recent advancements in imaging techniques for identifying difficult airways and explores the potential integration of AI into imaging-based predictive models.
Jain et al. (Thu,) studied this question.