Experimental validation of a multi-energy x-ray adapted scatter separation method

Area of Science:

• Medical Physics
• Radiological Imaging
• Photon Counting Detectors

Background:

• Energy-resolved photon counting detectors in radiography and CT enable material identification and quantification.
• Scattered radiation severely compromises X-ray image accuracy, reducing contrast and introducing bias/artifacts.
• Accurate material imaging necessitates high-quality, scatter-free X-ray images.

Purpose of the Study:

• To experimentally evaluate a partial attenuation spectral scatter separation approach (PASSSA) for multi-energy X-ray imaging.
• To assess the effectiveness of PASSSA in correcting scatter-induced artifacts and improving image quality.
• To validate PASSSA's performance against a reference beam-stop (BS) method.

Main Methods:

• Utilized a prototype X-ray system for radiographic acquisitions of an anthropomorphic thorax phantom.
• Acquired reference primary images using the beam-stop (BS) method.
• Applied the partial attenuation spectral scatter separation approach (PASSSA) for scatter correction.

Main Results:

• PASSSA-corrected attenuation images demonstrated significantly increased local contrast and improved internal structure visibility.
• A substantial reduction in scatter-induced bias was observed in corrected images compared to uncorrected ones.
• The PASSSA method showed good agreement with BS reference data, reducing normalized root-mean-square error (NRMSE) from 45% to approximately 5%.

Conclusions:

• The evaluated PASSSA technique effectively reduces scatter-induced bias in multi-energy X-ray imaging.
• PASSSA significantly enhances image quality, improving contrast and contour visibility for material identification.
• This scatter correction method shows promise for accurate material imaging applications using photon counting detectors.