RESEARCH ARTICLE Removal of noise in MRI images using a block difference-based filtering approach I. Nagarajan 1 | G.G. Lakshmi Priya 2 1 School of Computer Science and Engineering, VIT, Vellore, India 2 School of Information Technology and Engineering, VIT, Vellore, India Correspondence G.G. Lakshmi Priya, School of Information Technology and Engineering, VIT, Vellore, India Email: lakshmipriya.gg@vit.ac.in Abstract Magnetic resonance imaging (MRI) images are frequently sensitive to certain types of noises and artifacts. The denoising of MRI images is essential for improving visual quality and reliability of the quantitative analysis of diagnosis and treatment. In this article, a new block difference-based filtering method is proposed to denoise the MRI images. First, the normal MRI image is degraded by a certain percentage of noise. The block difference between the intensity of the normal and noisy MRI is computed, and then it is compared with the intensity of the blocks of the normal MRI image. Based on the comparison, the pixel weights are updated to each block of the denoised MRI image. Observational results are brought out on the BrainWeb and BraTS datasets and evaluated by performance metrics such as peak signal-to- noise ratio, structural similarity index measures, universal quality index, and root mean square error. The proposed method outperforms the existing denoising filter- ing techniques. KEYWORDS denoising, filtering, MRI, pixel similarity, Rician noise 1 | INTRODUCTION The magnetic resonance imaging (MRI) technique is one of the most popularly used imaging techniques for clinical diagnosis and treatment. MRI images are normally involved with low signal-to-noise ratios (SNRs). However, MRI images are degraded by certain noises and artifacts. There are several cases of noise occurrence in MRI images, which include thermal, Gaussian, and Rician noise. 1 These noises create image distortion and blurring, and also complicate the process of pulling up significant data for medical diagnosis. Thus, removing noise in MRI images is essential for better image visualization and promotes reliability of the associ- ated quantitative analysis. The process of denoising the image is a great challenge in the medical area. 2 The best approach to get good quality of MRI images is to average the multiple repeatedly acquired images. Nevertheless, it increases the acquisition time and is not suitable for application where quick methods are required. Nevertheless, practical implementation is not always possible because of the discomfort of patient and limitations of technical aspects. Some other significant aspect of noise is the thermally active electrons in the body of the patient, thus affecting the quality of MRI images. Hence, there is a necessity to develop an efficient method to get rid of the noise in the seized image. The denoising technique is generally classified into two types, which includes linear and nonlinear filtering tech- niques. 3 Furthermore, the linear filters are classified as spa- tial and temporal filters. In the case of linear filters, the noise is reduced by changing the image element value of weighted-mean neighborhood pixels and it creates a poor quality of image. 4 The nonlinear filter is processed within the filter windows. The neighboring pixels are organized along the basis of sample attributes of a windowpane. Con- sequently, it makes the right character of image. 5 The best solution for diagnosis made is a nonlinear filter rather than a Received: 11 July 2018 Revised: 4 June 2019 Accepted: 25 July 2019 DOI: 10.1002/ima.22361 Int J Imaging Syst Technol. 2019;1–13. wileyonlinelibrary.com/journal/ima © 2019 Wiley Periodicals, Inc. 1