minc_anlm

Adaptive non-local means denoising of MRI images.

minc_anlm [options] <source> <output>

DESCRIPTION

minc_anlm applies an adaptive non-local means (ANLM) denoising filter to a MINC volume. The method is based on Manjon, Coupe, Marti-Bonmati, Collins, and Robles (2010), “Adaptive Non-Local Means Denoising of MR Images With Spatially Varying Noise Levels” (Journal of Magnetic Resonance Imaging, 31(1):192-203).

The filter removes noise while preserving edges and structural details by averaging similar patches from a local search neighborhood. The adaptive variant automatically adjusts the filtering strength based on local noise estimates. A Rician noise correction mode is available for magnitude MRI data.

OPTIONS

--rician

Apply correction for Rician noise bias, appropriate for magnitude MRI data.

--search n

Set the search neighborhood radius in voxels.

--patch n

Set the patch radius in voxels for similarity comparisons.

--beta f

Adjust the smoothing weight parameter. Higher values produce more smoothing. Default: 1.

--mt n

Set the number of threads for parallel processing. Default: 1.

--verbose

Print progress information during processing.

--double

Write output in double-precision floating-point format.

--float

Write output in single-precision floating-point format.

--short

Write output in short integer format.

--byte

Write output in byte (unsigned char) format.

EXAMPLES

Denoise a T1-weighted MRI volume:

minc_anlm t1.mnc t1_denoised.mnc

Denoise with Rician noise correction and 4 threads:

minc_anlm --rician --mt 4 t1.mnc t1_denoised.mnc

Denoise with custom search and patch radii:

minc_anlm --search 3 --patch 1 --beta 0.8 t1.mnc t1_denoised.mnc

AUTHOR

Vladimir S. Fonov - McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University.

COPYRIGHTS

Copyright © 2011 by Vladimir S. Fonov

SEE ALSO

mincnlm, noise_estimate