A signal processing model of diagnostic x-ray scatter

Area of Science:

• Medical Physics
• Signal Processing
• Radiological Imaging

Background:

• X-ray scatter significantly impacts image quality in medical imaging.
• Existing models for scatter often rely on empirical data or complex simulations.

Purpose of the Study:

• To develop a signal processing-based model for X-ray scatter.
• To derive a closed-form expression for the scatter-to-primary ratio.

Main Methods:

• Developed a nonlinear filter (NLF) model for the scattering process.
• Derived the NLF point spread function (PSF) from single scattering principles (Compton scattering, X-ray attenuation).
• Characterized the PSF with approximations for geometric shape, volume, and width.

Main Results:

• Obtained a closed-form expression for scatter-to-primary ratio as a function of key parameters (phantom thickness, field size, energy, distances).
• Validated the NLF model against experimental measurements with constant thickness phantoms.
• Demonstrated good agreement between the model and measured data.

Conclusions:

• The signal processing approach provides a viable framework for modeling X-ray scatter.
• The NLF model successfully captures the functional dependence of scatter on various parameters.
• This model offers an alternative to empirical and Monte Carlo methods for scatter analysis.