Conventional reconstruction methods make a number of unrealistic assumptions about the imaging process in Atom Probe Tomography (APT), such as the shape of the sample apex. While holding relatively well for homogeneous materials, these assumptions result in spatial distortions and artifacts arising within heterogeneous materials (samples with internal structure). This can distort the internal structure of samples within reconstructions, and limit the technique's applicability. The severity of these distortions depend on both the material and sample geometry.
My research focuses on modelling the atom probe imaging process, and applying such models to improving reconstruction methods. By physically modelling the APT imaging process, many of the assumptions limiting current reconstruction methods vanish. Such a model-driven reconstruction method has the potential to correct many of the spatial distortions currently present within reconstructions.