Dual-Modal Safety Framework for Robotic-Assisted Bronchoscopy via Endoscopic Vision and Haptic Feedback

Area of Science:

• Medical Robotics
• Surgical Technology
• Endoscopy

Background:

• Flexible bronchoscopy lacks crucial distal haptic feedback, posing a risk of iatrogenic tissue trauma.
• Unobservable tissue damage can occur during endoluminal exploration in current teleoperated systems.

Purpose of the Study:

• To develop a robot-assisted bronchoscopy system with an integrated dual-modal safety framework.
• To enhance surgeon situational awareness and mitigate risks of iatrogenic trauma.
• To improve dexterity and safety in flexible bronchoscopy procedures.

Main Methods:

• A visual module using image moments for predictive collision avoidance alerts.
• A Transformer-based network estimating distal interaction torques for real-time haptic feedback.
• Integration of predictive visual warnings and real-time haptic feedback into a teleoperated system.

Main Results:

• Phantom validations confirmed the framework's effectiveness in tasks like fragile tissue handling and luminal exploration.
• Visual warnings significantly reduced excessive force application.
• The learning-based estimator provided reliable distal force perception.

Conclusions:

• The dual-modal safety architecture balances operational dexterity with safety in flexible bronchoscopy.
• The system enhances operator situational awareness, reducing the risk of iatrogenic trauma.
• Real-time haptic feedback and predictive visual alerts improve safety during robotic bronchoscopy.