Physical Modeling and Performance of Spatial-Spectral Filters for CT Material Decomposition.

Area of Science:

• Medical Imaging
• Physics
• Materials Science

Background:

• Spectral computed tomography (CT) enables material decomposition by analyzing multiple photon energy sensitivities in a single acquisition.
• Conventional spectral CT often relies on energy-discriminating detectors or multiple X-ray sources.

Purpose of the Study:

• To investigate a new spectral CT approach using spatially encoded spectra via K-edge filters.
• To model and analyze the impact of X-ray focal spot size and filter motion blur on performance.

Main Methods:

• A tiled pattern of K-edge filters is used to create spatially encoded spectral data.
• A spatial-spectral filter moves continuously relative to the X-ray source for improved sampling.
• A model-based material decomposition algorithm reconstructs material densities from sparse spectral channel projection data.

Main Results:

• Simulated focal spot widths from 0.2 mm to 4.0 mm were analyzed.
• Filter motion blur was simulated for translation speeds from 50 mm/s to 450 mm/s.
• Performance degradation with a 1.0 mm focal spot compared to 0.2 mm was under 15%, indicating feasibility with standard X-ray tubes.

Conclusions:

• The K-edge filter approach offers a viable alternative for spectral CT without specialized detectors.
• Increased filter motion speeds improve sampling and reduce error, outweighing spectral blur effects within practical ranges.