CT Image Reconstruction from Sparse Projections Using Adaptive TpV Regularization.

Area of Science:

• Medical Imaging
• Radiology
• Image Reconstruction

Background:

• Minimizing radiation dose in CT scans is crucial for patient safety.
• Sparse-projection reconstruction, using fewer X-ray views, offers a potential dose reduction strategy.
• Existing methods like total variation (TV) minimization have limitations.

Purpose of the Study:

• To develop a novel adaptive regularization technique for sparse-projection CT image reconstruction.
• To improve image quality and reduce artifacts in low-dose CT.
• To accelerate the iterative reconstruction process.

Main Methods:

• Introduction of a novel adaptive TpV (Total p-variation) regularization.
• Application of the Fast Iterative Shrinkage-Thresholding Algorithm (FISTA) for accelerated convergence.
• Numerical experiments comparing the proposed method with existing reconstruction techniques.

Main Results:

• The proposed adaptive TpV method effectively suppresses noise and artifacts.
• Superior preservation of structural information compared to other methods was observed.
• The FISTA technique accelerated iterative convergence.

Conclusions:

• The novel adaptive TpV regularization method is efficient for sparse-projection CT.
• This technique offers a promising approach for dose reduction in CT while preserving image quality.
• The method enhances diagnostic accuracy by improving image clarity and reducing artifacts.