Medical image super-resolution for remote medical diagnosis in smart city: A case study based on the new healthcare reform of China

Abstract

China is confronting with difficulty to tackle increased inequality in medical service between urban and rural areas. With the development of Internet of Things, telemedicine could provide remote clinic diagnosis in smart city via embedded systems for rural area. However, it is hard to acquire deserved resolution medical image since the limitation of hardware in rural area. Enhancing the image resolution is necessary for the medical images obtained in rural area. Anchor neighborhood regression based super-resolution is promising in enhancing the resolution of medical images. However, anchor neighborhood projections of ANR method are instability to cover variety styles of mapping relationships. In this respect, this paper propose a guided anchored neighborhood regression (GANR) based super-resolution method. The experiment results validates that our method has better reconstruction results in terms of both objective evaluation and subjective evaluation compare with previous methods.