Noise Analysis in MRI: Overview
Different methods have been reported in literature in order to estimate noise parameters out of a single MRI slice based on distinct models and situations. Those methods are extensively reviewed in the following chapters. However, when estimating noise out of a real acquisition, some considerations must be taken into account beyond the estimation method itself. In many occasions, an estimator cannot be directly applied over data, at a risk of miss-estimation of the parameters of noise. This chapter makes a profound analysis on how to estimate noise from MRI data in practical situations. The starting point is an example that raises the main issues concerning the noise estimation task. These issues will be deeply analyzed: the use of a noise model; the stationarity of the data; the use of the background in estimation; how the quantification of the data can alter the estimation or the use of multiple samples. Additionally, a practical scheme to effectively estimate noise out of MRI is proposed.