Reducing residual-motion artifacts in iterative 3D CBCT reconstruction in image-guided radiation therapy

Area of Science:

• Medical Physics
• Radiological Imaging
• Radiation Oncology

Background:

• Iterative cone-beam computed tomography (iCBCT) offers superior image quality over analytical methods like Feldkamp-Davis-Kress (FDK).
• Standard iCBCT utilizes penalized likelihood (PL) reconstruction, which is sensitive to physiological motion artifacts during data acquisition.

Purpose of the Study:

• To develop a computationally inexpensive extension of iCBCT that mitigates motion-related artifacts.
• To improve the robustness and image quality of iCBCT in the presence of patient motion during radiation therapy.

Main Methods:

• A modified iterative reconstruction approach was developed, altering the gradients of the PL function to prevent motion artifact generation.
• A motion simulation was created to generate CBCT projections of moving anatomy, with artifact-free images serving as ground truth.
• Quantitative analysis using contrast-to-noise ratio and power spectra of difference images assessed the impact of motion and the proposed modification.

Main Results:

• Patient abdominal wall motion during acquisition introduces low-frequency artifacts in standard iCBCT reconstructions.
• The proposed modification of PL effectively reduces these low-frequency artifacts.
• The modified iCBCT method suppresses motion-related artifacts in clinically relevant regions, enhancing motion resiliency without compromising PL advantages.

Conclusions:

• The modified iterative reconstruction method significantly enhances CBCT image quality for anatomies affected by residual motion.
• This technique offers improved motion artifact reduction in iCBCT for radiation therapy applications.