Reduction of Angularly-Varying-Data Truncation in C-Arm CBCT Imaging

Area of Science:

• Medical Imaging
• Computational Imaging
• Image Reconstruction

Background:

• C-arm cone-beam computed tomography (CBCT) provides 3D anatomical data for surgical and interventional procedures.
• Limited field-of-view (FOV) in CBCT causes data truncation, leading to artifacts that obscure details like tumors.
• Angularly-varying data truncation is a significant challenge in clinical CBCT applications.

Purpose of the Study:

• To develop and evaluate an optimization-based reconstruction method to reduce truncation artifacts in C-arm CBCT.
• To address the limitations of existing methods in handling angularly-varying data truncation.
• To improve the diagnostic utility of CBCT by enhancing image quality.

Main Methods:

• Formulated reconstruction as a constrained optimization program using truncated data.
• Incorporated a data-derivative-ℓ2-norm fidelity term to suppress artifacts.
• Tailored the Chambolle-Pock algorithm to solve the optimization problem.
• Collected data from physical phantoms and human subjects.

Main Results:

• The proposed optimization-based reconstruction effectively reduced image artifacts caused by angularly-varying data truncation.
• The method yielded images with improved clarity and reduced obscuration of low-contrast details.
• Demonstrated the feasibility of direct reconstruction from truncated data without pre-compensation.

Conclusions:

• Optimization-based reconstruction is a viable approach for mitigating truncation artifacts in C-arm CBCT.
• The tailored Chambolle-Pock algorithm effectively solves the proposed optimization problem.
• This technique has the potential to enhance the reliability and accuracy of CBCT in clinical practice.