Development of Advanced X-ray Micro-Tomography Helical Scan andTomographic Data Processing

Abstract: Micro-computed tomography (Micro-CT) is an imaging technique that allows for high-resolution,three-dimensional (3D) internal visualization of small objects in a non-destructive manner. However,a significant limitation arises when attempting to image elongated samples at higher resolutions andmagnifications due to the inherently small field of view (FOV). This thesis project introduces anapproach aimed at overcoming this limitation by combining helical scanning with the utilization ofthe open-source Astra Toolbox. The toolbox provides flexible scanning geometry parametrizationand facilitates tomographic reconstruction using the Feldkamp-Davis-Kress (FDK) algorithm.Helical data scanning represents a departure from traditional data acquisitions. Instead ofcapturing data at a fixed height, helical scanning acquires data by rotating and translating theimaged sample. This acquisition process extends the effective FOV, enabling the reconstruction oflarger samples. To illustrate this approach, a micro-CT helical scan was performed on a pencil asan elongated sample and compared with a conventional circular scan.While Micro-CT is a powerful imaging tool, it faces challenges when imaging weakly attenuating materials with hard X-rays. X-ray radiation possesses both amplitude and phase components,but conventional CT can only capture intensity information. This limitation often leads to poorcontrast in images of weakly absorbing samples. To address this issue, the project also designedpost-processing software to enhance CT image quality through phase retrieval. Phase informationprovides better contrast for small structures of weakly absorbing materials compared to intensityalone. The Paganin phase-retrieval algorithm is employed to retrieve this phase information, resulting in improved contrast-to-noise ratio (CNR) in tomographic reconstructions and subsequentsegmentation. As a result, the generated images exhibit sharper delineations and offer a morecomprehensive representation of the sample’s internal structure, especially after 3-D rendering.In summary, in this thesis, I designed the micro-CT procedure both from acquisition with anadvanced helical scanning technique and from a post-processing point of view through designingsoftware with a phase retrieval algorithm. The project has improved the conventional microCT intwo aspects, namely the field of view and the phase contrast.