Precise 3D/2D calibration between a RGB-D sensor and a C-arm fluoroscope

Area of Science:

• Medical Imaging
• Computer Vision
• Robotics

Background:

• Accurate calibration between RGB-D cameras and C-arm fluoroscopes is crucial for medical augmented and mixed reality applications.
• Existing methods face challenges in achieving precise registration for these integrated systems.

Purpose of the Study:

• To develop and evaluate an efficient 3D/2D calibration method for aligning RGB-D camera data with C-arm fluoroscope images.
• To improve the accuracy and reduce errors in the calibration process for medical imaging applications.

Main Methods:

• A custom calibration phantom combining traditional X-ray markers and a checkerboard plane was designed.
• The method localizes 2D markers in X-ray images and corresponding 3D points from RGB-D data.
• A projection matrix is calculated directly, bypassing simultaneous estimation of intrinsic and extrinsic parameters.

Main Results:

• Achieved a Root Mean Square (RMS) reprojection error of 0.5 mm, indicating high precision for surgical use.
• The proposed calibration method demonstrated superior accuracy compared to Tsai's method.
• Simulations confirmed that using a projection matrix reduces rotational angle errors.

Conclusions:

• A robust and effective 3D/2D calibration technique for integrating RGB-D cameras with C-arm fluoroscopes has been successfully developed and validated.
• The method offers significant improvements in accuracy and error reduction, making it suitable for advanced surgical navigation and augmented reality systems.