Cardiovascular diseases and cancer are leading causes of mortality, causing almost 30 million deaths annually. A very common technique for diagnosing these is x-ray computed tomography (CT), which reconstructs 3D images from x-ray measurements. The current CT technology is limited with respect to diagnostic quality and quantitative accuracy, which the emerging photon-counting CT technology can overcome with higher spatial resolution, lower noise, and improved material-selective imaging, especially in combination with deep-learning-based image reconstruction. We are developing deep-learning-based CT image reconstruction methods suited for generating highly accurate photon counting CT images together with maps of image uncertainty, by training deep neural networks able to map measured x-ray imaging data into images with as high accuracy and resolution as possible. Another important application of the combination of photon-counting CT with AI is to improve ultra-low-dose imaging, which is important for example for lung-cancer screening. During the previous NAISS compute allocations (2023/22-85, 2024/22-220, 2025/22-399,) we developed a deep-learning motion artifact correction method and evaluated it with simulated and measured photon-counting CT data. This improves quantitatively accurate tumour imaging since artifacts from the beating heart and breathing motion may otherwise lead to errors in measurements of tumours. We have also developed a proof-of-concept method investigating how dynamic CT images can be created from static ones with generative AI, which allows generating synthetic training data for motion-compensation methods. Finally, we have studied denoising methods and developed an initial denoising method for ultra-low-dose imaging. which can potentially help generating dose-efficient and reliable screening methods for cancer and cardiovascular disease. During the proposed project, we will further improve the cancer-imaging capabilities by extending this project in multiple potential directions: 1) We will continue the investigations into deep-learning based motion artifact correction with one-step joint image reconstruction and registration, with the aim of developing image reconstruction combined with simultaneous image reconstruction. This will bypass the need for reconstruction of an intermediate time series of CT images followed by registration. The ability to generate synthetic CT images with motion will increase the amount of training and testing data available. 2) We will also further develop our ultra-low-dose imaging technique for imaging cancer, which uses the available information optimally for doses as low as 10-1000 µSv, which can be used for lung-cancer screening. Next steps include optimizing the image quality, fine-tuning the method for actual photon-counting images and evaluating the diagnostic performance in human observer studies on real photon-counting CT images. The image data used with NAISS compute resources will consist of images of test objects and anonymized datasets from internet databases, i.e. no protected health information will be stored on NAISS servers. After training we will download the models, apply them to patient images acquired with a photon-counting CT prototype and evaluate their clinical usefulness locally.