DENOISING ABNORMAL MRI IMAGES

Abstract

Magnetic Resonance Image (MRI) is a medical imaging technique used in radiology to capture the anatomy and the physiological processes of the body. During this process, it gets affected by some unwanted noises like Gaussian noise, Salt and Pepper noise and Speckle noise. The removal of such noises from the actual MRI has been a hard nut to crack for researchers as it ultimately results in the formation of artifacts and causes blurred MR images. It is also considered a very vital step since it facilitates the analysis of the data of the image. There are many methods being used to remove or reduce noise.

This paper attempted to study two types of noise filtering techniques (Median filter and Wiener filter) on a noisy MRI intercepted by Salt and Pepper noise and Gaussian noise. Quality metrics such as MSE and PSNR were computed to evaluate the effectiveness of these two filters.

The results of the study noticed that it was the type of the filtering technique that decided the level removal or reduction of noise on these images. The researcher hopes that this result can be more useful in many medical diagnostics procedures and related applications.