RCAN based MRI super-resolution with applications

High-quality Magnetic Resonance Imaging (MRI) provides an aid to reliable diagnosis. By processing medical images, doctors are able to view pathological features more easily; however, since pathological features are usually small in the early stages, we would like to obtain medical images that are as clear as possible so that doctors can observe relevant features and even finer textures. Therefore, applying image super-resolution technology to medical images can reconstruct high-resolution medical images without increasing hardware equipment, which helps doctors make better diagnoses of patients' conditions. The use of deep learning methods can effectively reconstruct high-resolution images. In this paper, RCAN network is applied to the medical image dataset IXI and formed a control experiment with SAN, EDSR and T2Net network, and RCAN achieves good results both in terms of data metrics and in terms of recovering images for analysis.