Augmented reality-based surgical guidance for wrist arthroscopy with bone-shift compensation

Area of Science:

• Orthopedic Surgery
• Medical Imaging
• Computer-Assisted Surgery

Background:

• Intraoperative joint positioning differs from preoperative imaging due to surgical motion, leading to inaccurate surgical targeting.
• Existing methods lack real-time adaptation to intraoperative joint changes during wrist arthroscopy.

Purpose of the Study:

• To develop and validate a real-time augmented reality (AR)-based surgical guidance system for wrist arthroscopy.
• To implement a bone-shift compensation model to address intraoperative joint motion and improve surgical accuracy.

Main Methods:

• A novel bone-shift compensation method using noninvasive fiducial markers and virtual link models was developed.
• The system measures and compensates for wrist traction during surgery, updating virtual bone models in real-time.
• Camera calibration and patient-specific registration were performed for AR projection onto arthroscopic images.

Main Results:

• The bone-shift compensation method estimated wrist bone position during traction with a 1.4 mm margin of error.
• Significant reduction in target point error was observed for multiple wrist bones (e.g., 74.8% for trapezoid) after compensation.
• Augmented reality display expanded the arthroscope's field of view and improved visualization of anatomical structures.

Conclusions:

• The proposed method successfully mitigates AR errors caused by wrist traction during arthroscopy.
• Accurate AR visualization of concealed bones and expanded field of view enhance surgical guidance.
• The bone-shift compensation technique is adaptable to other joints (knee, shoulder) and can utilize noninvasive markers for skin movement.