Toward an End-to-End Calibration for Mobile C-Arm in Combination with a Depth Sensor for Surgical Augmented Reality...

Area of Science:

• Medical Imaging
• Computer-Assisted Surgery
• Robotics

Background:

• C-arm X-ray imaging is crucial for orthopedic surgery guidance.
• Augmented Reality (AR) enhances surgical procedures by overlaying C-arm images.
• Accurate calibration of C-arm systems for AR remains a significant challenge due to imaging limitations.

Purpose of the Study:

• To propose and evaluate a novel, accurate, and robust calibration procedure for surgical AR systems utilizing C-arm X-ray and RGB-D sensors.
• To address the limitations of existing calibration methods, including C-arm instability and narrow field of view.
• To ensure the calibration procedure is consistent with standard surgical protocols.

Main Methods:

• A new calibration procedure involving C-arm imaging system calibration and 3D/2D calibration of an RGB-D camera and C-arm system.
• Application of bundle adjustment equations with a 3D multi-modal phantom (Lego).
• Utilizing a virtual detector to compensate for C-arm movement effects.

Main Results:

• Achieved a mean overlay error of 1.03 mm for visualizing X-ray images on 3D data.
• Demonstrated promising accuracy and improved flexibility and robustness compared to existing methods.
• Successfully compensated for C-arm movement effects on calibration parameters, improving overlay error with non-zero C-arm rotation.

Conclusions:

• The proposed calibration procedure offers a reliable and accurate solution for surgical AR applications using C-arm and RGB-D sensors.
• The method enhances the flexibility and robustness of C-arm calibration for AR surgeries.
• The technique effectively mitigates the impact of C-arm movement, leading to improved overlay accuracy.