Sparse view cone beam X-ray luminescence tomography based on truncated singular value decomposition

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Computational Imaging

Background:

• Cone beam X-ray luminescence computed tomography (CB-XLCT) offers fast imaging but suffers from ill-conditioned inverse problems, leading to poor image quality, especially for multiple targets.
• Accurate and fast imaging of multi-targets is crucial for in vivo applications, but current CB-XLCT methods struggle to achieve this.

Purpose of the Study:

• To develop a fast and effective reconstruction method for multi-target CB-XLCT imaging.
• To improve the image quality of CB-XLCT for resolving multiple targets with varying distances.

Main Methods:

• A truncated singular value decomposition (TSVD) based sparse view CB-XLCT reconstruction method was proposed.
• TSVD was used to orthogonalize the system weight matrix, enabling the application of compressed sensing (CS) based L1-norm for reconstruction from limited projection views.

Main Results:

• The proposed method effectively resolved two targets with different edge-to-edge distances (EEDs) in numerical simulations and phantom experiments.
• Significant improvements in localization accuracy, target shape, image contrast, and spatial resolution were observed for multi-target imaging.

Conclusions:

• The TSVD-based sparse view CB-XLCT method offers a viable solution for fast and high-quality multi-target imaging.
• This advancement holds promise for enhancing in vivo applications requiring precise visualization of multiple targets.