Denoising

The program vdenoise removes noise from MRI data using wavelet filtering. Wavelet filters may produce pseduo-Gibbs artefacts. Such artefacts should be reduced using a second program called vsrad. Denoising has been tested on structural images (T1-weighted,Flash) and diffusion weighted images. In principle, it should also be usable on fMRI data, but it has not yet been tested in this context.

The filter strength is controlled by the parameter ‘-level’. Reducing the value of ‘-level’ reduces the filter effect and also the possibility of artefacts. The type of wavelet may be specified by the parameter ‘-wavelet’. The ‘haar’ wavelet is a good choice. Here is a typical calling sequence.

vdenoise -in input.v -out tmp.v -level 0.5
vsrad -in tmp.v -out output.v -iter 5

Reference: Lohmann et al (2010). Magn Res Med. 64(1), pp:15-22.

Parameters of ‘vdenoise’

-help

Usage information.

-in

Input file.

-out

Output file.

-level

Filter strength. Default: 0.5.

-wavelet

Wavelet function to be used. Default: haar. Options: [haar | daub4 | daub6 | daub8 | bspline103 | bspline202].