Principal Component Analysis Based Dynamic Cone Beam X-Ray Luminescence Computed Tomography: A Feasibility Study.

Area of Science:

• Biomedical Imaging
• Medical Physics
• Radiology

Background:

• Cone beam X-ray luminescence computed tomography (CB-XLCT) is valuable for studying physiological and pathological processes in small animals.
• Accurately capturing dynamic bio-distributions of probes, especially in adjacent targets, remains a challenge for conventional CB-XLCT.

Purpose of the Study:

• To develop a novel 4D temporal-spatial reconstruction method for dynamic CB-XLCT.
• To improve the resolution and accuracy of imaging dynamic probe bio-distributions in small animals.
• To enhance the ability to distinguish and track probes in adjacent targets.

Main Methods:

• A 4D temporal-spatial reconstruction method utilizing principal component analysis (PCA) in the projection space was developed.
• Projections were compressed to reduce noise, followed by temporal PCA to decouple the 4D problem into 3D problems.
• Reconstruction was performed using restarted Tikhonov regularization in the PCA domain, incorporating principal components reflecting target dynamics.

Main Results:

• The proposed PCA-based method effectively resolves multiple targets and recovers dynamic bio-distributions.
• Numerical simulations and phantom experiments validated the method's ability to achieve high computation efficiency.
• The technique demonstrated improved accuracy in distinguishing dynamic behaviors of probes in adjacent targets.

Conclusions:

• The developed 4D reconstruction method offers a significant advancement for dynamic CB-XLCT imaging.
• This approach provides new feasibility for in vivo imaging of dynamic probe bio-distributions in small animal studies.
• The method enhances the study of physiological and pathological processes by improving the visualization of probe dynamics.