Detection of difficult airway using deep learning

Aguilar, Kevin; Alférez, Germán H.; Aguilar, Christian

doi:10.1007/s00138-019-01055-3

Cited by 10 publications

(9 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This suggests a better sensitivity to the combined technique for finding difficult airways without sacrificing too much specificity. In addition, the predictive value (AUC) in difficult laryngoscopy (0.62) and intubation (0.79) presented certain value but were not ideal when compared with recent evidences (15,22). More standardized teaching strategies for Chinese ED practitioners who intubate may help improve the utility of these airway assessments.…”

Section: Discussionmentioning

confidence: 99%

Incidence and diagnostic validity of difficult airway in emergency departments in China: a cross-sectional survey

Dai¹,

Walline²,

Wu³

et al. 2023

J Thorac Dis

View full text Add to dashboard Cite

Background:The difficult airway is a topic of concern for any physician performing intubation. Severe adverse events which can happen because of an unexpected difficult airway emphasize the importance of prediction. This study aimed to investigate the incidence of difficult airways in emergency departments in mainland China and ascertain the predictive value of common assessment formulas.Methods: This study was a multicentral, cross-sectional investigation of patients intubated in emergency departments in mainland China between March 1-30, 2021. Frequencies and percentages were calculated for patient characteristics. Diagnostic values were analyzed by sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, the area under the curve, and corresponding P values.Results: 985 patients in 41 hospitals were enrolled in the study. Elderly male patients occupied most of the intubated patients. The difficult face mask ventilation rate (8.2%) was higher, but the difficult glottic exposure (6.8%) and intubation (2.1%) rates were low, while there was widespread use of enhanced visualization equipment (80.6%). Four combined emergency applicable evaluation methods showed a certain predictive value for the difficult airway with a sensitivity (0.71; 95% CI: 0.47-0.89), specificity (0.79; 95% CI: 0.76-0.81), and the area under the curve (0.75; 95% CI: 0.64-0.87, P<0.01).Conclusions: Despite an insufficient predictive ability for current difficult airway evaluation methods, there were relatively low rates of difficult airways during ED intubations in mainland China. Continuing focus on the difficult airway problems and optimizing airway assessment are required.

show abstract

Section: Discussionmentioning

confidence: 99%

Incidence and diagnostic validity of difficult airway in emergency departments in China: a cross-sectional survey

Dai¹,

Walline²,

Wu³

et al. 2023

J Thorac Dis

View full text Add to dashboard Cite

show abstract

“…Previous studies using machine learning algorithms to predict difficult laryngoscopy, difficult airway or difficult intubation have reported high predictive power [27][28][29], but they all required indicators that should be evaluated by clinicians. In contrast, our method does not require any human evaluation.…”

Section: Discussionmentioning

confidence: 99%

Deep‐learning model associating lateral cervical radiographic features withCormack–Lehanegrade 3 or 4 glottic view

et al. 2022

View full text Add to dashboard Cite

Summary Unanticipated difficult laryngoscopy is associated with serious airway‐related complications. We aimed to develop and test a convolutional neural network‐based deep‐learning model that uses lateral cervical spine radiographs to predict Cormack–Lehane grade 3 or 4 direct laryngoscopy views of the glottis. We analysed the radiographs of 5939 thyroid surgery patients at our hospital, 253 (4%) of whom had grade 3 or 4 glottic views. We used 10 randomly sampled datasets to train a model. We compared the new model with six similar models (VGG, ResNet, Xception, ResNext, DenseNet and SENet). The Brier score (95%CI) of the new model, 0.023 (0.021–0.025), was lower (‘better’) than the other models: VGG, 0.034 (0.034–0.035); ResNet, 0.033 (0.033–0.035); Xception, 0.032 (0.031–0.033); ResNext, 0.033 (0.032–0.033); DenseNet, 0.030 (0.029–0.032); SENet, 0.031 (0.029–0.032), all p < 0.001. We calculated mean (95%CI) of the new model for: R2, 0.428 (0.388–0.468); mean squared error, 0.023 (0.021–0.025); mean absolute error, 0.048 (0.046–0.049); balanced accuracy, 0.713 (0.684–0.742); and area under the receiver operating characteristic curve, 0.965 (0.962–0.969). Radiographic features around the hyoid bone, pharynx and cervical spine were associated with grade 3 and 4 glottic views.

show abstract

“…MobileNetV2 [58] -This architecture uses an inverted residual structure with linear bottlenecks, and is specifically designed to demand limited computational resources. Examples of its use include facial recognition [37], MMT score estimation [59], and segmentation of the airway anatomy [60].…”

Section: Deep Learningmentioning

confidence: 99%

Automated location of orofacial landmarks to characterize airway morphology in anaesthesia via deep convolutional neural networks

García-García

Lee

Mendoza-Garcés

et al. 2023

Computer Methods and Programs in Biomedicine

View full text Add to dashboard Cite

Detection of difficult airway using deep learning

Cited by 10 publications

References 18 publications

Incidence and diagnostic validity of difficult airway in emergency departments in China: a cross-sectional survey

Incidence and diagnostic validity of difficult airway in emergency departments in China: a cross-sectional survey

Deep‐learning model associating lateral cervical radiographic features withCormack–Lehanegrade 3 or 4 glottic view

Automated location of orofacial landmarks to characterize airway morphology in anaesthesia via deep convolutional neural networks

Contact Info

Product

Resources

About