Estimation and removal of noise from Magnetic Resonance Images (MRI) is an active area of research. Noise remains one of the main causes of quality deterioration in MRI and is a subject in a large number of papers in the MRI literature. Consideration of how noise affects the true signal is important for proper interpretation and analysis of MR images. This work mainly deals with the estimation of noise and the underlying true signal from both single and multiple coil acquired magnetic resonance images.