Iterative image-domain ring artifact removal in cone-beam CT.

Area of Science:

• Medical Imaging
• Image Processing
• Radiology

Background:

• Ring artifacts in cone beam computed tomography (CBCT) images, caused by flat-panel detector pixel gain variations, degrade image quality.
• These artifacts manifest as structured non-uniformities, impacting diagnostic accuracy and treatment planning.

Purpose of the Study:

• To propose a general method for effective ring artifact removal in CBCT images.
• To enhance image quality and preserve structural details in CBCT scans.

Main Methods:

• A novel method utilizing a polar coordinate system transforms ring artifacts into stripe artifacts.
• Relative total variation is employed for image smoothing, followed by iterative subtraction and compensation steps to isolate and remove artifacts.
• The iterative process refines artifact removal by minimizing differences in extracted ring artifacts.

Main Results:

• Evaluated using phantoms and clinical data, the method significantly improved spatial uniformity by 1.68 times.
• Structural similarity index increased from 87.12% to 95.50% in simulated data.
• Clinical data demonstrated high efficiency in artifact removal, preserving image structure and detail.

Conclusions:

• The proposed iterative ring artifact removal method is practical and effective for CBCT imaging.
• This technique offers significant improvements for CBCT-guided radiation therapy by enhancing image quality.
• The method successfully removes artifacts while maintaining essential image features.