Study of the Noise Reduction Algorithm with Median Modified Wiener Filter for T2-weighted Magnetic Resonance Brain Images

The purpose of this study was to confirm the utility of the median modified Weiner filter (MMWF) noise reduction algorithm in T-2-weighted brain MR images. We acquired brain MR images using both real experiment and simulation, and performed comparative evaluation with conventional noise reduction algorithms through calculation of factors for the noise level and similarity. As a result, evaluation of the noise level and similarity showed the most improved in the image, which the MMWF noise reduction algorithm was applied. Moreover, additional experiment was conducted using real MR device and water phantom to more clearly prove the efficiency of the MMWF noise reduction algorithm. The noise level and intensity profile results derived from the real T-2-weighted MR image proved the effectiveness of the MMWF noise reduction algorithm. In conclusion, the proposed MMWF noise reduction algorithm demonstrated a very promising performance for improving the image quality in T-2-weighted MR image.