Super resolution image reconstruction in parallel magnetic resonance imaging

In clinical applications, images with high resolution are often desired and required which may provide more details for doctors to make precise diagnosis. In this paper, an approach is proposed to increase image resolution of parallel magnetic resonance imaging. Since different receiver coils have different sensitivity profiles, different receiver channel models are constructed to map the original image information to low resolution images of different channels. Based on these models, the degradation function of every low resolution image can be obtained to compute the high resolution image iteratively using the well known super resolution approach. An in-vivo experiment is also provided to illustrate the feasibility and robustness of the proposed approach.