Sci-Sat AM(1): Imaging-05: Analytical scatter estimation for cone-beam computed tomography.

Area of Science:

• Medical Physics
• Image Reconstruction
• Computational Imaging

Background:

• High scatter-to-primary ratio in cone-beam computed tomography (CBCT) degrades image quality, causing artifacts like cupping and shading.
• Reduced contrast and increased noise in CBCT images hinder accurate diagnosis and high-resolution imaging.
• Effective scatter reduction methods are crucial for advancing CBCT implementation in clinical settings.

Purpose of the Study:

• To develop and validate a computational method for scatter estimation and compensation in CBCT.
• To integrate an accurate scatter estimator into a statistical reconstruction algorithm for improved image quality.
• To assess the feasibility of an analytical approach for scatter estimation in CBCT.

Main Methods:

• Developed an analytical method for single scatter estimation utilizing Klein-Nishina cross-sections.
• Compared analytical scatter estimates with high-count EGSnrc Monte Carlo simulations for validation.
• Extended phantom studies from small homogeneous phantoms to larger, more clinically relevant heterogeneous phantoms.

Main Results:

• Analytical scatter estimates demonstrated favorable agreement with Monte Carlo simulation results.
• The method was validated on larger, more complex phantoms simulating breast tissue with contrast inserts.
• Computational acceleration was achieved using parallel processing on a High-Performance Computing network.

Conclusions:

• The developed analytical method provides a computationally tractable approach for estimating single scatter in CBCT.
• This method shows potential for artifact reduction and image quality enhancement in CBCT.
• The findings support the integration of this scatter estimation technique into statistical reconstruction algorithms for clinical CBCT systems.