Nonlocal maximum likelihood estimation method for denoising multiple-coil magnetic resonance images Jeny Rajan a,* , Jelle Veraart a , Johan Van Audekerke b , Marleen Verhoye b , Jan Sijbers a a IBBT-Vision Lab, Dept. of Physics, University of Antwerp, Belgium b Bio-Imaging Lab, Dept. of Biomedical Sciences, University of Antwerp, Belgium Abstract Effective denoising is vital for proper analysis and accurate quantitative mea- surements from magnetic resonance (MR) images. Even though many meth- ods were proposed to denoise MR images, only few deal with the estimation of true signal from MR images acquired with phased array coils. If the mag- nitude data from phased array coils are reconstructed as the root sum of squares, in the absence of noise correlations and subsampling, the data is assumed to follow a non central-χ distribution. However, when the k -space is subsampled to increase the acquisition speed (as in GRAPPA like meth- ods), noise becomes spatially varying. In this note, we propose a method to denoise multiple-coil acquired MR images. Both the non central-χ distribu- tion and the spatially varying nature of the noise is taken into account in the proposed method. Experiments were conducted on both simulated and real data sets to validate and to demonstrate the effectiveness of the proposed method. Keywords: MRI; Noise; Denoising; NLML; Non central chi distribution 1. Introduction Stochastic noise is one of the main causes of quality deterioration in Mag- netic Resonance Images (MRI) and, hence, estimation and removal of noise remains an active area of research. Consideration of how noise affects the * Corresponding author Email address: jeny.rajan@ua.ac.be (Jeny Rajan) Published in Magnetic Resonance Imaging July 30, 2012