An optimization-based method for geometrical calibration in cone-beam CT without dedicated phantoms.

Area of Science:

• Medical Imaging
• Computed Tomography
• Image Reconstruction

Background:

• Geometrical misalignments in cone-beam CT scanners degrade image quality.
• Accurate calibration is crucial for reliable tomographic reconstruction.
• Existing methods often require specific phantoms or object information.

Purpose of the Study:

• To develop a novel, model-independent method for geometrical misalignment determination in cone-beam CT.
• To enable scanner calibration using only projection data of a generic object.
• To assess the accuracy and limitations of the proposed calibration technique.

Main Methods:

• A cost function is defined based on projection data, dependent on misalignment parameters.
• Scanner calibration is achieved by minimizing this cost function using optimization techniques.
• The method is validated through numerical simulations and applied to micro-CT data.

Main Results:

• The method accurately determines transversal detector shift and skew.
• Detector slant can also be determined.
• Parameters with negligible impact on image quality, like detector tilt, are identified as indeterminable by this method.

Conclusions:

• The proposed method offers a robust and accurate approach for cone-beam CT scanner calibration.
• It is applicable to generic objects, reducing the need for specialized phantoms.
• The technique shows promise for various cone-beam imaging modalities.