Deep Learning Techniques for CT Image Denoising and Resolution Enhancement

In recent years, deep learning (DL) has significantly impacted computed tomography (CT) image reconstruction. This chapter explores two critical areas where DL has substantially improved CT image quality: denoising and spatial resolution enhancement. We provide a comprehensive review of various DL-based CT image denoising algorithms, ranging from conventional regression models to cutting-edge diffusion-based generative models. Additionally, we examine DLs remarkable contributions to image super resolution, showcasing how sharper and more accurate images can be produced reliably under limited hardware conditions. The chapter presents representative algorithms and illustrative image examples to demonstrate these advancements. We seek to offer an in-depth examination of deep learning applications in CT image reconstruction, highlighting the substantial progress in the field and its potential for ongoing innovation in medical imaging.