DSP Based rician noise removal in MR images using gradient projection algorithm

Rician noise in Magnetic Resonance images affects the image contrast, signal dependent bias and signal to noise ratio value. The combined Gabor filter with Gradient projection filter was developed to remove the Rician noise in Magnetic Resonance images and compared its performance with the state-of-the-art filtering methods. The algorithm was developed in the TMS320C6713DSK, which exhibited improvements in the image quality and signal to noise ratio value. Further, it reduces the noise component and preserves the image features with less execution time. For testing the noise removal algorithm, two different levels of Rician noise were added with the raw MR images and removed using the popular and the proposed method. The performance of the proposed method was compared with the popular methods through the estimation of quantitative metrics and visual inspection.