Interior Reconstruction Using the Truncated Hilbert Transform via Singular Value Decomposition.

Area of Science:

• Medical Imaging
• Computational Imaging
• Applied Mathematics

Background:

• Current exact local computed tomography (CT) reconstruction relies on iterative Projection Onto Convex Sets (POCS) methods.
• The POCS method, while effective, is computationally intensive and time-consuming.
• There is a need for faster and more efficient algorithms for localized CT reconstruction.

Purpose of the Study:

• To develop and evaluate a novel method for exact local CT reconstruction using singular value decomposition (SVD).
• To compare the computational efficiency and reconstruction quality of the SVD method against the traditional POCS approach.
• To demonstrate the feasibility of SVD-based localized CT for pre-clinical and clinical applications.

Main Methods:

• Reconstruction of the object function using the truncated Hilbert transform (THT) via singular value decomposition (SVD).
• Implementation of the proposed SVD algorithm.
• Performance evaluation through numerical simulations and comparison with the iterative POCS method.

Main Results:

• The SVD-based approach achieves reconstruction speeds two orders of magnitude faster than the iterative POCS method.
• The SVD method produces excellent region-of-interest (ROI) reconstructions, enabling previously impossible imaging.
• Numerical simulations confirm the feasibility and efficiency of the proposed SVD technique.

Conclusions:

• Singular value decomposition (SVD) offers a significantly faster and effective alternative for exact local computed tomography (CT) reconstruction.
• The SVD method facilitates localized pre-clinical and clinical CT, opening new research avenues in exact local image reconstruction.
• This research demonstrates the potential of SVD for advancing the field of medical imaging and image reconstruction.