Autonomous Frameless Spine Registration and Robot-Assisted Surgery

Area of Science:

• Robotics and biomechanics
• Medical imaging and image analysis
• Surgical navigation systems

Background:

• Accurate patient registration is crucial for robot-assisted spine surgery.
• Current methods involve invasive frames or complex intraoperative imaging, increasing risks and costs.
• A need exists for efficient, non-invasive registration techniques.

Purpose of the Study:

• To develop and evaluate a novel non-invasive, contactless spine registration system.
• To enable accurate patient registration for robot-assisted spine surgery.
• To reduce radiation exposure and procedural complexity associated with current methods.

Main Methods:

• A system integrating a skin-affixed, radiopaque mask and a robot-mounted monocular camera was developed.
• The system uses live camera feeds and advanced algorithms for real-time mask tracking and pose estimation.
• Phantom experiments were conducted to assess registration accuracy and pedicle-screw targeting.

Main Results:

• The system demonstrated millimeter accuracy in pedicle-screw path localization on spine phantoms.
• Registration was completed in under 4 minutes per vertebra.
• The system effectively accounts for patient positioning changes between CT scanning and the operating theater.

Conclusions:

• The developed system provides an objective and accurate reference for image-guided screw placement in spine surgery.
• It facilitates frame-free patient registration for both manual and robotic procedures.
• The system lays the groundwork for future clinical translation, pending cadaveric and pilot studies.