Improved L1/L2 minimization algorithm for segmental limited-angle CT reconstruction.

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Imaging

Background:

• Computed tomography (CT) is vital in various fields but faces limitations like high radiation doses and long scan times.
• Segmental limited-angle (SLA) CT sampling strategies aim to mitigate these issues but often result in shading artifacts in reconstructed images.

Purpose of the Study:

• To address shading artifacts in SLA CT reconstruction.
• To enhance image quality in limited-angle CT by introducing a novel regularization technique.

Main Methods:

• A reconstruction model incorporating the L1/L2 ratio of image gradients as a regularization term for SLA CT.
• Development of an improved L1/L2 minimization algorithm, including linearization of the data fidelity term and Fast Fourier Transform (FFT) acceleration.
• Application of the alternating direction method for image reconstruction.

Main Results:

• The proposed L1/L2 method demonstrated superior performance compared to existing methods in numerical simulations and real CT data experiments.
• Effective preservation of image structures and fine details in reconstructed images.
• Significant reduction in shading artifacts inherent to SLA CT.

Conclusions:

• The L1/L2 regularization technique offers a promising solution for improving image quality in segmental limited-angle CT.
• The developed algorithm efficiently reconstructs high-quality CT images with preserved details and reduced artifacts.
• This method has the potential to enhance the applicability of CT in various domains by improving image fidelity and reducing scan burdens.