Purpose: Our group has shown that incorporating models of scatter into statistical iterative image reconstructions for quantitative multirow CT (MRCT) improves accuracy. However, experimental measurements and Monte Carlo simulations gave different answers to the basic question, how does the scatter-to-primary ratio (SPR) change with increasing kVp? This issue is also unclear in the literature, with the majority of papers showing SPR decreasing with kVp, while the remaining are split between increasing and constant SPR. Recognizing the complexity of this situation, Kalender argued in 1981 that â€œIt is absolutely necessary to specify all experimental parametersâ€?. We propose a simple first-collision physical model that predicts the dependence of SPR on x-ray energy and other key system parameters.
Methods: The Hangartner model of scatter was modified to include the differential cross sections for the coherent and incoherent scattering processes. Planar water objects were studied, with cross-sections using form factors accounting for molecular and intermolecular forces. The 3D scanner geometry included typical 1D anti-scatter grids (ASG) and MRCT energy-integrating detectors. For every studied condition, at each object point within a detector field-of-view, the scatter angle and flux were calculated. The SPR were studied as a function of x-ray energy, collimation height, object thickness, and ASG numerical aperture.
Results: For 1D ASG grids with acceptance angles close to modern MRCT detector designs (~1 degree), SPR increases with increasing kVp. For wider acceptance angles (>5 degrees), SPR decreases with increasing kVp. This can be attributed to the narrowing of the coherent-scattering forward profile with increasing energy, which increases flux into a narrow-angle ASG, while a wide-angle ASG includes the majority of the profile at all energies.
Conclusion: The dependence of SPR on x-ray energy depends on the acceptance aperture of the detector, increasing with energy for narrow-angle MRCT systems and decreasing for wide-angle CBCT systems.
Funding Support, Disclosures, and Conflict of Interest: This study was supported by NIH 1R01CA212638