Using Computerized Facial Image Analysis to Predict Difficulty of Endotracheal Intubation Compared to Standard Airway Exams

What this Trial Is About

Researchers are evaluating a computer algorithm designed to predict how easy or difficult it is to perform endotracheal intubation by analyzing digital photographs of a patients face from three different angles. This study compares the algorithms predictions with traditional airway assessment techniques to improve accuracy and patient safety during intubation procedures. The study includes patients who are easy or difficult to intubate, based on specific surgical criteria and anesthetic records. Participants are divided into groups for model development and validation, including those identified as easy or difficult to intubate and a group without prior intubation difficulty classification. Photographs of the head and neck are taken from the front, left, and right sides, which are then analyzed using facial structure software to create face models. The study also allows for intubation attempts using either a Macintosh or Miller laryngoscope blade, with the aim to assess facial features predictive of difficulty with different blades. During the study, participants undergo photographic imaging of their head and neck, and their intubation outcomes are recorded at surgery. Researchers will compare the algorithms ability to predict intubation difficulty against conventional tests by analyzing the images and surgical outcomes over approximately two years. The study tracks whether the algorithm can statistically differentiate between easy and difficult intubation cases and tests the reproducibility of the developed models to improve clinical acceptance and patient safety.

Facial Analysis to Classify Difficult Intubation