A novel weighted total difference based image reconstruction algorithm for few-view computed tomography.

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Imaging

Background:

• Computed tomography (CT) imaging requires high-quality images from limited projection data to minimize patient radiation dose.
• Limited-view CT reconstruction often results in severe artifacts and blurred object edges.
• Compressed sensing algorithms are gaining attention for limited-view CT reconstruction.

Purpose of the Study:

• To develop a novel iterative reconstruction algorithm for few-view CT.
• To eliminate streak artifacts and preserve edge structure information in CT images.
• To demonstrate the superior performance of the proposed algorithm over conventional methods.

Main Methods:

• Proposed a novel iterative reconstruction algorithm based on weighted total difference (WTD) minimization.
• The WTD measure enforces sparsity and directional continuity in the gradient domain.
• Employed a soft-threshold filtering approach to solve the WTD-based few-view CT reconstruction model.

Main Results:

• The WTD algorithm significantly outperformed the total difference (TD)-based algorithm in numerical experiments.
• Achieved over 60% gains in RMSE, NRMSD, and NMAD, and over 10% in PSNR with 40 projections.
• Demonstrated over 15% gains in RMSE, NRMSD, and NMAD, and over 4% in PSNR with noisy projections.

Conclusions:

• The proposed WTD algorithm effectively suppresses streak artifacts in few-view CT reconstruction.
• The algorithm successfully preserves crucial edge structure information.
• The WTD-based approach offers superior performance for low-dose CT imaging applications.