Brain MR image super-resolution reconstruction via sparse representation and non-local similarity regularisation

Abstract

In magnetic resonance imaging, image resolution is limited by several factors such as hardware constraints or physical considerations. In many cases, the acquired images have to be magnified to match a specific resolution. This paper presents a new super-resolution reconstruction algorithm to generate a high-resolution version of a low-resolution brain MR image. The proposed approach uses a multi-scale first- and second-order derivative analysis to estimate the missing high-frequency information and integrates sparse representation and non-local similarity regularisation into a unified L1norm minimisation framework. Extensive experiments on brain MR image super-resolution validate that the proposed method achieves much better results than many state-of-the-art algorithms in terms of both quantitative measures and visual perception.