Researchers from the University of Shanghai for Science and Technology and Fudan University’s Eye & ENT Hospital systematically review advances in AI-assisted tracheal intubation robotics and anatomical recognition algorithms. They analyze developmental stages from integrated to intelligent designs, evaluate robotic systems such as KIS and REALITI, and discuss AI techniques like CNNs and visual servo control. The review outlines challenges and clinical implications for improving intubation success rates and operational efficiency.
Key points
- Kepler Intubation System (KIS) achieved a 91% clinical first-pass success rate with an average intubation time of 57 s.
- REALITI automated robot uses a 2-DOF continuum endoscope with visual servo control for glottis navigation in mannequin trials.
- YOLO-U-Net cascade algorithm delivers >95% IoU in epiglottis and vocal cord segmentation at 10+ FPS on simulated airway images.
Why it matters: Integrating AI and robotics in airway management promises safer, faster intubations, reducing complications and resource constraints in critical care settings.
Q&A
- What is tracheal intubation?
- How do robotic arms improve intubation precision?
- What is visual servo control in airway robotics?
- How do CNN-based models recognize airway structures?
