Noise Reduction in MRI Images

Abstract

Real world signals usually contain departures from the ideal signal that would be produced by our model of the signal production process. Such departures are referred to as noise. Noise arises as a result of unmodelled processes going on in the production and capture of the real signal. It is not part of the ideal signal and may be caused by a wide range of sources, e.g. variations in the detector sensitivity, environmental variations, the discrete nature of radiation, transmission or quantization errors, etc. It is also possible to treat irrelevant scene details as if they are image noise. The characteristics of noise depend on its source, as does the operator which best reduces its effects. In this study, we have tried to analyze the different types of noises present in MRI images and to filter these noises using different digital filters. In the given figure, we added poisson, salt & pepper, speckle, Gaussian, additive Gaussian and multiplicative Gaussian noise. In the present study, we quantitatively establish the use of various parameters which helps in analysing the filter performance that are otherwise difficult to determine by other classical methods of image processing. This study investigates which type of filter removes or reduces which noise properly and if combination of filters are used then which type of filters give the desired results. The present study focuses on max, min, median, cumulative mean, bilateral and Gaussian filter. The parameters which determine the filter performance are PSNR, WPSNR, Correlation, SNR.