Related Experiment Video
Updated: May 22, 2026
06:09
Measuring 3D In-vivo Shoulder Kinematics using Biplanar Videoradiography
Published on: March 12, 2021
An EPID based method for performing high accuracy calibration between an optical external marker tracking device and
Zachary Grelewicz1, Hyejoo Kang, Rodney D Wiersma
1Department of Radiation and Cellular Oncology, The University of Chicago, Chicago, IL 60637, USA. zachg@uchicago.edu
Medical Physics
|May 8, 2012
Summary
A new semiautomated method accurately calibrates 3D optical cameras to linear accelerators (LINACs) using an electronic portal imaging device (EPID). This approach improves spatial registration for radiation therapy, exceeding current accuracy standards.
Area of Science:
- Medical Physics
- Radiation Oncology
- Image Guidance
Background:
- External 3D optical tracking cameras are increasingly used in modern radiation therapy.
- Accurate camera-to-linear accelerator (LINAC) reference frame calibration is crucial for precise treatment delivery.
- Current manual calibration methods are prone to errors.
Purpose of the Study:
- To develop a semiautomated, quantitative calibration method for camera-to-LINAC reference frames.
- To eliminate errors associated with manual alignment procedures.
- To utilize only an electronic portal imaging device (EPID) for calibration.
Main Methods:
- A phantom with seven IR reflective BBs was imaged using a 3D stereoscopic IR imager and an MV EPID.
- Simulated annealing optimized BB locations in the LINAC frame based on optical positions and EPID images.
- Singular value decomposition calculated the transformation matrix between camera and LINAC frames.
Main Results:
- Simulated data yielded BB location accuracy with a Root Mean Square Error (RMSE) of 0.23 mm.
- Experimental data from MV imager yielded BB location accuracy within 0.26 mm.
- Accuracy was dependent on the chosen optimization function.
Conclusions:
- The developed method achieves highly accurate spatial registration between external 3D imaging and LINAC frames.
- Experimental results surpassed the 1 mm accuracy threshold.
- The method meets and exceeds the AAPM TG-142 standard of ≤2 mm positioning accuracy.
