Comparison of Endotracheal Intubation Performance Using Video Laryngoscopy With and Without AI-Based Visual...

Area of Science:

• Medical technology
• Anesthesiology
• Artificial intelligence in medicine

Background:

• Video laryngoscopy (VL) has a steep learning curve for novice practitioners.
• Improving VL training is crucial for patient safety and procedural efficiency.

Purpose of the Study:

• To evaluate the effectiveness of AI-based visual assistance in reducing the learning curve for video laryngoscopy.
• To assess the impact of AI-enhanced VL (AI-VL) on intubation performance in inexperienced volunteers.

Main Methods:

• Twenty volunteers without prior intubation experience were recruited.
• Participants were randomized into two groups: standard VL and AI-VL.
• Each participant performed 10 consecutive intubations, with performance metrics recorded.

Main Results:

• The AI-VL group demonstrated a trend towards a steeper learning curve for intubation time (P > .05).
• The AI system acted as a virtual instructor, offering real-time visual feedback.
• This feedback assisted in maintaining optimal Cormack-Lehane (C-L) views during intubation.

Conclusions:

• AI-based visual assistance in VL shows promise as a training tool.
• AI-VL may help accelerate skill acquisition and improve view maintenance in novice users.
• Further research is warranted to confirm the impact of AI-VL on clinical outcomes.