X-Ray Computed Tomography (CT) is a non-destructive imaging technique that computes cross-sectional images from a set of measurements. This process is used in many industrial application, and there is room for adoption of this technique in the forestry industry. At present, the use of this advanced technique is limited by the price of scanners and their acquisition speed. This study aims at developping scanners tailored for scanning logs, as well as the bespoke machine-learning techniques to post-process the CT images (segmentation of knots, wood density estimation, geometric properties determination). Thus, we have a need of computational resources to first simulate the scanning process and to train machine-learning algorithms to reconstruct and post-process the images. Then, we will need to run the algorithms on real data acquired by scanners located at Luleå university. Working on real data from scanners is an incredible opportunity to explore the adaptation of the reconstruction algorithms to the downstream, post-processing tasks, which can potentially have spillover effects in other industries were automatic post-processing of CT images is implemented, such as in medical imaging.