Modeling Shift-Variant X-Ray Focal Spot Blur for High-Resolution Flat-Panel Cone-Beam CT

Area of Science:

• Medical Imaging
• Radiology
• Computational Imaging

Background:

• Flat-panel cone-beam CT (CBCT) is used in high-resolution medical imaging.
• Geometric magnification in CBCT can improve resolution but introduces focal-spot blur.
• Existing reconstruction methods may not adequately account for blur variations.

Purpose of the Study:

• To develop and validate a shift-variant focal-spot blur model for CBCT.
• To improve image quality and spatial resolution in CBCT reconstructions.
• To address limitations in current CBCT system design and performance.

Main Methods:

• Developed a shift-variant focal-spot blur model.
• Incorporated the model into a model-based iterative reconstruction algorithm.
• Applied the algorithm to simulated trabecular bone data and physical CBCT experiments.

Main Results:

• Traditional CBCT reconstruction without blur modeling showed shift-variant resolution dependent on acquisition protocols and anode angles.
• Focal spot blur was characterized in physical CBCT experiments.
• Image quality significantly improved with the shift-variant model compared to no blur or shift-invariant blur models.

Conclusions:

• A shift-variant focal-spot blur model can significantly enhance CBCT image quality.
• This approach offers a potential solution to overcome spatial resolution and system design limitations in CBCT.
• The developed model improves reconstruction accuracy by accounting for spatially varying blur characteristics.