Prior Image Guided Ring Artifact Correction for Photon-Counting Detector Data in Hybrid Spectral CT

Area of Science:

• Medical Imaging
• X-ray Detector Technology
• Image Processing

Background:

• Photon-counting detectors (PCDs) offer superior spatial resolution and dose efficiency in X-ray imaging.
• Non-uniformity in PCDs is a significant challenge, causing ring artifacts in computed tomography (CT) images.
• Existing methods for artifact correction require further development for advanced detector systems.

Purpose of the Study:

• To propose and validate a novel image-domain post-processing method for correcting ring artifacts in PCD-reconstructed CT images.
• To leverage the strengths of hybrid spectral CT systems by integrating data from both PCD and energy integration detectors (EIDs).
• To enhance the diagnostic quality of CT images acquired with PCD technology.

Main Methods:

• Developed an image-domain post-processing technique for ring artifact correction.
• Utilized ring artifact-free EID-reconstructed images as prior information within an optimization framework.
• Incorporated a weighted optimization model to preserve boundary details and mitigate artifacts from prior image blurring.

Main Results:

• The proposed method successfully corrected ring artifacts in PCD-reconstructed images.
• Validation was performed using both simulated and real datasets, confirming the method's effectiveness.
• The technique demonstrated efficient and effective artifact removal, improving image quality.

Conclusions:

• The developed post-processing method offers a practical solution for addressing ring artifacts in PCD-based hybrid spectral CT imaging.
• This approach enhances the clinical utility of advanced photon-counting detector technology.
• The findings support the integration of PCDs in future CT systems.