Adding noise to real MRI data

1. How to add Rician noise?
2. What is Rician noise?
3. What distribution is used to model noise variation in MRI?

How to add Rician noise?

One way to add Rician noise is via a complex temporary image: Simply add normally distributed (i. e., Gaussian) complex noise to the image, and then calculate the magnitude of the complex image data. The result is Rician-distributed image data (and, in particular, Rayleigh-distributed noise in the background).

What is Rician noise?

The distribution of measured MR pixel intensities in the presence of noise is known to be Rician, and the width of this dis- tribution is directly related to the Gaussian noise on the measured real and imaginary signals. It is the pixel magnitude values that follow the Rician distribution, not the noise.

What distribution is used to model noise variation in MRI?

It is common practice to assume the noise in magnitude MRI images is described by a Gaussian distribution. The power of the noise is then often estimated from the standard deviation of the pixel signal intensity in an image region with no NMR signal.