An analytical form of ring artifact correction for computed tomography based on directional gradient domain...

Area of Science:

• Medical Imaging
• Image Processing
• Computational Science

Background:

• Ring artifacts are a common issue in Computed Tomography (CT) imaging, often stemming from detector imperfections or filters.
• Existing physics-based correction methods struggle to completely eliminate these artifacts, necessitating robust image-domain solutions.
• The presence of ring artifacts can compromise diagnostic accuracy and treatment planning in CT scans.

Purpose of the Study:

• To develop an effective and robust method for removing ring artifacts from reconstructed CT images.
• To preserve intricate image details during the artifact correction process.
• To enhance the clinical utility of CT imaging through improved image quality.

Main Methods:

• The proposed method converts CT images to polar coordinates, transforming artifacts into stripes.
• Relative Total Variation is employed to extract the image's structural information.
• Directional Gradient Domain Optimization (DGDO) is introduced to restore details using gradient and structural information, followed by an analytical algorithm for minimization.

Main Results:

• The method successfully corrected ring artifacts in both synthetic and real-world CT images.
• Experiments demonstrated superior visual quality compared to previous artifact removal techniques.
• The approach effectively preserved intricate details within the CT images.

Conclusions:

• The developed method provides an analytical solution for ring artifact removal in CT images while preserving essential details.
• This technique holds significant practical value in the medical field for improving CT image quality.
• By enhancing image quality and reducing diagnostic challenges, the method contributes to improved patient care.