Dual spectral limited-angle CT imaging regularized by edge-preserving diffusion and smoothing

Area of Science:

• Medical Imaging
• Materials Science
• Computer Vision

Background:

• Limited-angle CT is crucial for nondestructive inspection of planar objects.
• X-ray beam hardening poses challenges for high-absorption materials like metal.
• Existing CT methods struggle with artifacts in such scenarios.

Purpose of the Study:

• To develop a dual spectral limited-angle CT scan method.
• To propose an image reconstruction algorithm mitigating beam hardening and metal artifacts.
• To improve image quality in limited-angle CT for challenging materials.

Main Methods:

• Utilized a dual spectral limited-angle CT scan approach.
• Developed a reconstruction algorithm considering X-ray polychromatic properties.
• Applied the method to simulated and real PCB (Printed Circuit Board) data.

Main Results:

• The proposed algorithm effectively removed beam hardening artifacts.
• Metal artifacts were significantly reduced in the reconstructed images.
• Limited-angle artifacts were also notably depressed.
• Demonstrated effectiveness on both simulated and real-world PCB data.

Conclusions:

• Dual spectral limited-angle CT is effective for nondestructive inspection.
• The developed algorithm successfully overcomes limitations of existing methods.
• This technique enhances imaging of planar objects with high-absorption materials.