Superiorization-based multi-energy CT image reconstruction

Area of Science:

• Medical Imaging
• Computational Science

Background:

• Constrained optimization problems require robust solutions.
• Multi-energy CT (CT) image reconstruction benefits from advanced algorithms.
• The prior rank, intensity, and sparsity model (PRISM) offers regularization for CT reconstruction.

Purpose of the Study:

• To propose a superiorized simultaneous algebraic reconstruction technique (SART) algorithm for multi-energy CT image reconstruction using the PRISM model.
• To compare the performance of the proposed Superiorized-SART algorithm against the Split-Bregman algorithm.

Main Methods:

• Development of a superiorized SART algorithm incorporating the PRISM regularization model.
• Numerical experiments comparing the Superiorized-SART algorithm with the Split-Bregman algorithm for multi-energy CT image reconstruction.

Main Results:

• Both the Superiorized-SART and Split-Bregman algorithms produced high-quality CT images.
• Both algorithms demonstrated effectiveness in reducing image noise and artifacts.
• The superiorization approach proved efficient and robust for this application.

Conclusions:

• The proposed Superiorized-SART algorithm is a viable and effective method for multi-energy CT image reconstruction.
• Superiorization is a promising technique for improving CT image quality by reducing noise and artifacts.