Image Reconstruction Using Deep Learning
Aneeta Christopher, R. Hari Kishan, P. V. Sudeep · 2022
In computer vision, low-level vision tasks for recovering damaged images to high-quality images are referred to as image reconstruction. Image reconstruction is utilized in the medical imaging domain to get higher-quality images for clinical application at a low cost and risk to patients. Because there are several types of degradation that can cause considerable harm to images during the acquisition, it is an inevitable pre-processing task for many image processing applications. Several techniques for image reconstruction have already been investigated. Recently, deep learning has made significant strides in the field of image restoration and enhancement. This chapter investigates different image reconstruction approaches based on deep neural networks (DNNs) such as autoencoders (AEs), convolutional neural networks (CNNs), and generative adversarial networks (GANs). In addition, we present the usefulness of deep-learning architectures in applications such as image denoising, inpainting, and super-resolution. Finally, we discuss deep-learning-based image reconstruction techniques for medical imaging modalities such as magnetic resonance imaging (MRI), computed tomography (CT), medical ultrasound, optical coherence tomography (OCT), positron emission tomography (PET), and single-photon emission computerized tomography (SPECT).