Constrained regularized reconstruction of X-ray-DPCI tomograms with weighted-norm.

Area of Science:

• Medical Imaging
• Computational Imaging
• Physics

Background:

• X-ray differential phase-contrast imaging (DPCI) offers enhanced sensitivity for material differentiation.
• Reconstruction algorithms are crucial for obtaining high-quality DPCI data.
• Existing methods may suffer from artifacts and slow convergence.

Purpose of the Study:

• To introduce a novel and efficient reconstruction algorithm for DPCI.
• To improve the quality of reconstructed images by reducing artifacts.
• To demonstrate the algorithm's potential for in-vivo and medical imaging.

Main Methods:

• A variational formulation with a weighted data term.
• A variable-splitting scheme for fast convergence and artifact reduction.
• Incorporation of higher-order total-variation regularization and prior information (support, positivity).

Main Results:

• The proposed algorithm significantly reduces reconstruction artifacts.
• The use of higher-order regularization and prior information enhances image quality.
• Successful reconstruction experiments with real data were performed.

Conclusions:

• The new DPCI reconstruction algorithm offers improved performance.
• The method shows promise for various medical imaging applications.
• This approach facilitates better visualization and analysis in DPCI.