A separable quadratic surrogate total variation minimization algorithm for accelerating accurate CT reconstruction...

Area of Science:

• Medical Imaging
• Computational Science

Background:

• Total Variation Minimization (TVM) is crucial for accurate CT reconstruction with limited data.
• Existing TVM algorithms are computationally intensive, hindering practical application.

Purpose of the Study:

• To develop an accelerated algorithm for high-quality CT reconstruction from few-views and limited-angle data.
• To overcome the time-consuming optimization limitations of previous TVM methods.

Main Methods:

• Proposed a novel optimization algorithm leveraging the optimization transfer principle.
• Utilized TVM as a regularization term and a squared difference fidelity term in the cost function.
• Introduced a Separable Quadratic Surrogate (SQS) function based on the convexity of the terms to approximate the cost function.

Main Results:

• The algorithm successfully reconstructed high-quality CT images from limited view data, validated by simulations and real experimental data.
• Achieved reconstruction speeds over 10 times faster than previous algorithms.
• Demonstrated minimal differences in image quality compared to existing methods.

Conclusions:

• The novel algorithm significantly accelerates CT reconstruction from few-views and limited-angle data.
• Offers a more efficient solution for accurate CT imaging in scenarios with data constraints.