The reference implementation for the Non Local Spatial and Angular Matching (NLSAM) denoising algorithm for diffusion MRI.

The easiest way is to go grab a release, in which case the downloaded zip file contains everything you need (no python installation required, you can use it straight away without installing anything else). After extracting the zip file, start a terminal/command line prompt (start button, then type cmd + enter on windows) and navigate to where you extracted the binaries.

Since the tools are command line only, double-clicking it will open and immediately close a dos-like window, hence the need for opening a command line prompt.

If you would like to look at the code and modify it, you can build it from source by following the instructions. You can also download the datasets used in the paper over here.

Once installed, there are two main scripts, the stabilization algorithm and the NLSAM algorithm itself. The first one allows you to transform the data to Gaussian distributed signals if your dataset is Rician or Noncentral chi distributed.

A simple example call for the stabilization would be

stabilizer dwi.nii.gz dwi_stab.nii.gz 1 sigma.nii.gz -m mask.nii.gz --bvals bvals --bvecs bvecs 

and for the NLSAM denoising

nlsam dwi_stab.nii.gz dwi_nlsam.nii.gz 5 bvals bvecs sigma.nii.gz -m mask.nii.gz

You can find a detailed usage example and assorted dataset to try out in the example folder.

You will need to have at least numpy, scipy, nibabel, dipy, cython, cython-gsl and spams installed with python 2.7. Fortunately, the setup.py will take care of installing everything you need.

• On Debian/Ubuntu, you will need some development headers which can be installed with

sudo apt-get install build-essential libgsl0-dev python-dev libopenblas-dev libopenblas-base liblapack-dev

Of course feel free to use your favorite blas/lapack implementation (such as intel MKL), but I got 5x faster runtimes out of openblas vs atlas for NLSAM just by switching libraries.

• On Windows and Mac OSX, it will be easier to grab a python distribution which includes everything such as Anaconda. Additionally, grab a build of spams for windows here if you don't want to build it.
• Additionally, you might need Visual C++ Compiler for python 2.7 if you encounter the Unable to find vsvarsall.bat error when installing nlsam on Windows.

If you have a working python setup already, doing

pip install https://github.com/samuelstjean/nlsam/archive/master.zip --user --process-dependency-links

should give you everything you need.

You can also clone it locally and then build the files with

git clone https://github.com/samuelstjean/nlsam.git
cd nlsam
python setup.py build_ext -i

Don't forget to add the path where you cloned everything to your PYTHONPATH.

If you need help or would like more information, don't hesitate to drop me a line at firstname@isi.uu.nl, where of course firstname needs to be replace with samuel.

St-Jean, S., P. Coupé, and M. Descoteaux. "Non Local Spatial and Angular Matching : Enabling higher spatial resolution diffusion MRI datasets through adaptive denoising." Medical Image Analysis, 2016. DOI URL

As the main solver I use (spams) is GPL licensed and the stabilization script uses the GNU GSL library, the NLSAM main codebase is also licensed under the GPL v3, see the file LICENSE for more information.

If you would like to reuse parts of this work under another project/license, feel free to drop me an email and I will gladly re-license the files you need as MIT/BSD/whatever else.