A unified scatter rejection and correction method for cone beam computed tomography.

Area of Science:

• Medical Physics
• Radiological Imaging
• Cone Beam CT (CBCT)

Background:

• Scattered radiation significantly degrades image quality in flat panel detector-based CBCT.
• Existing 2D antiscatter grids effectively reduce scatter but residual scatter and grid-induced artifacts hinder quantitative imaging.
• Residual scatter poses a challenge for advanced techniques like dual-energy imaging in CBCT.

Purpose of the Study:

• To introduce a novel method for robust scatter rejection and residual scatter correction using 2D antiscatter grids.
• To transform 2D grids from scatter rejection devices into scatter measurement devices.
• To enhance the quantitative accuracy and reduce artifacts in CBCT imaging.

Main Methods:

• Developed Grid-based Scatter Sampling (GSS), where radiopaque grid septa create periodic shadows on the detector.
• Correlated modulated pixel intensity in septal shadows with local scatter intensity to measure residual scatter.
• Implemented GSS experimentally and evaluated its performance using simulations and phantom measurements.

Main Results:

• GSS method significantly improved CT number accuracy, reducing HU nonuniformity from 65 HU to 30 HU in pelvis phantoms.
• Residual scatter correction via GSS suppressed grid-induced ring artifacts, leading to a 41% reduction in noise.
• Spatial resolution, as indicated by the modulation transfer function (MTF), was preserved after artifact suppression.

Conclusions:

• The GSS method effectively utilizes 2D grids for both scatter measurement and correction in CBCT.
• This approach substantially improves quantitative accuracy, narrowing the image quality gap between CBCT and multi-detector CT.
• GSS successfully suppressed projection line artifacts and associated CBCT ring artifacts without compromising spatial resolution.