Conventional 3D X-ray microscopes distinguish different structures, by capturing differences in how these structures absorb X-rays. Recently, researchers have shown that modelling the more "wave-like" properties of X-rays can enhance the resolving power of 3D X-ray microscopes. In this new "phase contrast" X-ray microscopy, modelling the diffraction of the X-rays between the sample and detector, allows us to capture information about structures in the sample that are entirely invisible to conventional methods.

 

In the attached image of microstructure in wood, the dark and bright fringes result from un-modelled X-ray refraction.  As these fringes result from wave-optics rather than sample features, they confound automated analysis. Models of X-ray diffraction can be inverted (analytically, and/or using neural fields) to turn these fringes into valuable information about the internal structure of samples.

 

The ANU CTLab is a 3D X-ray microscopy facility, servicing industry and academic scientists in a variety of fields. The X-ray microscopes used at the CTLab utilise a unique "high cone-angle" scanning geometry, that allows us to inspect in greater detail the wave-like behaviour of X-rays within samples. This project will involve building a unified model of several theoretically-complex X-ray behaviours within the sample, drawing from statistical and wave optics: spatial partial-coherence, refraction, and spectral interactions. The student will then apply this model to improve imaging capabilities at the ANU CTLab.

• Mathematics (Fourier transforms, Multivariable calculus, Linear algebra, Statistics)
• Physics (Optics)
• Some experience in programming, preferably python.