Feasibility of similarity coefficient map in improving quality of magnetic resonance images of spleen

The purpose of this paper is to investigate the feasibility of similarity coefficient map (SCM) in improving the quality of MR images of spleen by improving signal to noise ratio (SNR) and contrast to noise ratio (CNR), revealing fine structures buried in noise, and integrating the information of each individual image in a series together into one image. To achieve these goals, in vivo experiments were conducted based on the data sets acquired using 12-echo T₂* weighted imaging. Preliminary results have demonstrated that compared with the original images, SCM images can (1) improve SNR by 14.50% to 104.44% mainly depending on the SNR in original images, (2) improve CNR by 82.95% to 103.52%, mainly depending on the choice of reference tissue, (3) reveal fine structures hidden in the original images such as tiny veins, and (4) offer a new type of contrast by integrating information of all original images. Moreover, It suggests that the higher the SNR of the reference, the higher the CNR of the resulting SCM images. In conclusion, SCM is a powerful post processing technique for a series of MR images of spleen with varying acquisition parameters. It can improve SNR and CNR, increase the amount of anatomical information by revealing previous invisible fine structures, and produce images with a new type of contrast that integrates information of all original images together.