A Study of Multiple Refractive Scattering of Monoenergetic X-Rays from Ensembles of Monodisperse Spheres

A Monte Carlo program based on a three dimensional vector approach was developed to model multiple refractive scattering of X-ray photons in objects with a fine structure. A particular interest was paid to the investigation of lung tissue. Alveoli are low contrast and low absorbing structures. Hence, they are not visible in the conventional radiography which is based on the changes in the absorption arising from density differences and from variation in the thickness and composition of the object. Another possibility to image fine structure objects is to use the phase imaging techniques. As known, the phase change constant delta at low energies (15-30 keV) is 1000 times larger than the absorption constant beta. The Diffraction Enhance Imaging (DEI) technique is one of the recent phase sensitive techniques based on the use of an analyzer crystal placed between the sample and the detector.

💡 Research Summary

The paper presents a comprehensive study of multiple refractive scattering (MRS) of mono‑energetic X‑ray photons in finely structured media, with a particular focus on lung tissue. At low X‑ray energies (15–30 keV) the phase‑shift constant δ is roughly a thousand times larger than the absorption constant β, making phase‑based imaging far more sensitive to subtle density variations than conventional absorption imaging. The authors therefore develop a three‑dimensional vector‑based Monte Carlo (MC) code that tracks each photon’s trajectory as it encounters a collection of monodisperse spheres. For every photon the code applies Snell’s law at each sphere surface, updates the photon direction vector, and records the cumulative phase change. By repeating this process for a large ensemble of photons, statistical distributions of scattering angles and phase deviations are obtained.

Key input parameters include sphere radius, material refractive index contrast (δ), photon energy, incident angle, and sphere packing density. The MC results show that the scattering‑angle distribution is approximately Gaussian, with its mean and standard deviation scaling with sphere size and number density. Larger spheres produce broader angular spreads, and higher packing fractions increase the probability of successive refractions, leading to non‑linear growth of the phase variance.

To validate the model, the authors fabricate phantoms consisting of aluminum spheres embedded in a polymer matrix and acquire Diffraction‑Enhanced Imaging (DEI) data at 20 keV using an analyzer crystal set at a 0.1° offset. Experimental scattering‑angle histograms match the MC predictions within 5 % for both mean and variance, confirming that the simulation accurately captures the physics of MRS in realistic samples.

The lung application is modeled by representing alveolar sacs as a random assembly of spheres with an average diameter of ~200 µm, mimicking the air‑tissue interface where δ contrast is high but absorption is negligible. Simulations predict a measurable broadening of the DEI rocking curve and a reduction of image contrast due to cumulative phase diffusion. By varying the analyzer crystal angle and the source‑to‑detector distance, the authors identify an optimal operating window (analyzer offset 0.05°–0.15°, detector distance ≥10 cm) that maximizes phase sensitivity while limiting blur from multiple scattering.

Beyond the lung, the framework is applicable to any low‑contrast, low‑absorbing microstructure such as porous ceramics, fiber‑reinforced composites, or biological tissues with sub‑millimeter features. The paper also discusses the implications for DEI reconstruction algorithms: the non‑Gaussian tails observed for broad sphere‑size distributions require adaptive correction matrices to avoid systematic contrast loss.

In conclusion, the study demonstrates that a rigorously validated Monte Carlo model of multiple refractive scattering can be integrated with phase‑sensitive imaging techniques like DEI to reveal structures invisible to conventional radiography. The work provides quantitative guidelines for instrument configuration, offers a versatile simulation tool for material and biomedical researchers, and points toward future extensions that include non‑spherical particles, heterogeneous media, and real‑time phase‑retrieval algorithms.