Pre-processing of Multimodal MR Images using NLM and Histogram Equalization

Abstract

Magnetic resonance imaging has emerged as one of the maxima broadly used and flexible scientific imaging modalities today. While at the beginning of this technique the research was fundamentally qualitative and relied exclusively on observation of images, currently several technological developments have driven it possible to derive quantitative measurements from these data. However, MR images are normally affected by several types of artifacts which must be minimized before a quantitative biomarker assessment pipeline is applied. This paper proposes preprocessing to the magnetic resonance (MR) images to enhance the standard of the image and to ease the next processing and analysis. The proposed method uses Non-Local Means (NLM) for noise removal and Histogram Equalization (HE) for enhancing the contrast of Magnetic resonance images for the preprocessing.