Facial Image Analysis for Fully Automatic Prediction of Difficult Endotracheal Intubation

Abstract

Abstract-Goal: Difficult tracheal intubation is a major cause of anesthesia related injuries with potential life threatening complications. Detection and anticipation of difficult airway in the preoperative period is thus crucial for the patients' safety. We propose an automatic face analysis approach to detect morphological traits related to difficult intubation and improve its prediction. Methods: For this purpose, we have collected a database of 970 patients including photos, videos and ground truth data. Specific statistical face models have been learned using the faces in our database providing an automated parametrization of the facial morphology. The most discriminative morphological features are selected through the importance ranking provided by the random forest algorithm. The random forest approach has also been used to train a classifier on these selected features. We compare a threshold tuning method based on class prior with two methods which learn an optimal threshold on a training set for tackling the inherent imbalanced nature of the database. Results: Our fully-automated method achieves an AUC of 81.0% in a simplified experimental setup where only easy and difficult patients are considered. A further validation on the entire database has proven that our method is applicable for real-world difficult intubation prediction, with AUC = 77.9%. Conclusion:The system performance is in line with the state-ofthe-art medical diagnosis, based on ratings provided by trained anesthesiologists, whose assessment is guided by an extensive set of criteria. Significance: We present the first completely automatic and non-invasive difficult intubation detection system that is suitable for use in clinical settings.

The role of artificial intelligence in preoperative medicine

“…Existing methods typically use combinations of medical history and physical examination to estimate the likelihood of a challenging airway management 18–21. Some AI methods use objective measurements, such as body mass index and thyromental distance, as features for prediction, while others have also described computerized facial analysis and photographs 22–25. The latter approaches may prove to be beneficial as many preoperative clinic evaluations have transitioned either partially or fully to telemedicine formats, spurred by the COVID-19 pandemic 26–28.…”

“…[18][19][20][21] Some AI methods use objective measurements, such as body mass index and thyromental distance, as features for prediction, while others have also described computerized facial analysis and photographs. [22][23][24][25] The latter approaches may prove to be beneficial as many preoperative clinic evaluations have transitioned either partially or fully to telemedicine formats, spurred by the COVID-19 pandemic. [26][27][28] Automatically generating an alert for a difficult airway can allow the preoperative physician to discuss potential airway management techniques with patients, including risks and benefits, and notify the intraoperative clinicians to prepare appropriate airway equipment.…”

Artificial intelligence, machine learning and the pediatric airway

“…The role of machine learning as a means to supplement the physical examination in predicting ease of intubation has been explored by Cuendet et al They proposed an automatic face analysis approach to detect morphological features related to difficult intubation. Using a database of 970 patient images/videos and their corresponding ease of intubation when subsequently undergoing a general anesthetic, statistical face models were created using automated parametrization.…”

Neural network prediction of difficult tracheal intubation risk by using the patient’s face image

“…G.L. Cuendet et al применили полностью авто матический подход к анализу изображений лица для идентификации морфологических признаков, связанных с ТИТ [15]. Для этой цели была собрана база данных из 970 пациентов, включающая фотографии, видеоролики и другие данные.…”