[Cone beam CT image iterative reconstruction based on Split-Bregman method].

Area of Science:

• Medical Imaging
• Computational Imaging
• Image Reconstruction

Background:

• Sparse-view cone beam CT (CBCT) is crucial for reducing radiation dose and scan time.
• Traditional reconstruction methods struggle with limited projection data, leading to image artifacts and reduced quality.
• Effective reconstruction algorithms are needed to overcome these limitations.

Purpose of the Study:

• To develop and evaluate a new iterative reconstruction method for sparse-view CBCT.
• To improve the accuracy and efficiency of CBCT image reconstruction using tight frame regularization.
• To enable dose reduction and faster reconstruction speeds.

Main Methods:

• A split-Bregman iterative method was employed for image reconstruction.
• Tight frame regularization was incorporated into the objective function, leveraging compressed sensing principles.
• The minimization problem was addressed by transforming L1 regularization to L2 and using the conjugate-gradient method, with intermediate variable updates via Bregman's method.

Main Results:

• The proposed method demonstrated significant advantages in image quality compared to existing techniques.
• Reconstruction time was notably reduced, indicating improved computational efficiency.
• Experimental results with digital and physical phantoms validated the approach's applicability and performance.

Conclusions:

• The novel split-Bregman method with tight frame regularization accurately reconstructs CBCT images from limited data.
• This approach facilitates lower X-ray doses and faster calculations compared to the POCS method.
• The method offers a promising solution for efficient and high-quality sparse-view CBCT reconstruction.