Fast noniterative calibration of an external motion tracking device

Area of Science:

• Medical Imaging
• Biomedical Engineering

Background:

• Prospective motion correction in magnetic resonance (MR) scans often relies on external tracking devices like cameras.
• Accurate translation of external tracking data into the MR scanner's reference frame requires precise knowledge of the camera's pose relative to the scanner.

Purpose of the Study:

• To describe a novel, rapid, accurate, and non-iterative technique for determining the position of an external tracking device relative to the MR reference frame.
• To enable precise hand-eye calibration for various motion tracking setups in magnetic resonance imaging (MRI).

Main Methods:

• The method involves imaging a sparse object capable of simultaneous tracking of rigid body transformations in both the MRI machine and external tracking device reference frames.
• A dual quaternion algorithm is employed to solve the calibration problem efficiently.

Main Results:

• The technique achieves high accuracy (0.2 mm, ~1/10 voxel size) in measuring large motions within the MRI reference frame.
• Effective camera calibration is accomplished with fewer than six measurements.
• Iterative application and motion correction feedback can further refine accuracy.

Conclusions:

• Independent tracking of movements across two reference frames provides an analytical solution to the hand-eye calibration problem in MRI.
• This method offers a significant advancement for motion tracking setups in MRI, improving accuracy and efficiency.