Singularity containerized version of the NODDI toolbox. NODDI matlab toolbox is a MatLab based toolbox developed by Prof Gary Zang for fitting NODDI data. Since the tool requires MatLab, it cannot be deplyed on High-Performance Computing (HPC) clusters that do not have MatLab installed on it. This NODDI singularity container provides an alternative option to run NODDI without MATLAB and can be readily deployed in most HCP clusters. For this container, we are using the single thread option of the NODDI matlab toolbox because the Parallel Computing Toolbox implementation requires an additional Matlab licence. The single thread version takes more time than the parallel version. However, this delay could be compensated when deployed in HPC where many images could be processed in parallel. (Eg. NODDI processing using Parallel Computing Toolbox to fit NODDI data with 220454 voxels in local machine using 8 parallel core takes ~ 3 hours. Thats 300 Hours for 100 subjets. The single thread version takes ~18 hours per subject. However, 100 subjects could be launched parallely in the HCP clusters.)
- Siya Sherif,
s.sherif_at_uliege.be
- Martin Grignard,
mar.grignard_at_uliege.be
- Christophe Phillips,
c.phillips_at_uliege.be
The container can be built locally on a machine where Singularity is available
singularity build noddi-singularity.sif singularity-ss-noddi.def
The singularity image can be dowloaded from zenodo or sylabs
or
# Pull with Singularity
$ singularity pull --arch amd64 library://sher_ss/default/noddi-singularity:1.0.1
# Pull by unique ID (reproducible even if tags change)
$ singularity pull library://sher_ss/default/noddi-singularity:sha256.6f5e08a8c425199cb1c1a48cef7cd03c9d39b59d3a862ad2ba780792d26d73ac
cite us, if you use the container
Zhang, H., Schneider, T., Wheeler-Kingshott, C. A., & Alexander, D. C. (2012). NODDI: practical in vivo neurite orientation dispersion and density imaging of the human brain. NeuroImage, 61(4), 1000–1016.
Sherif, S., Grignard, M., & Phillips, C. (July 12, 2021). noddi-singularity Zenodo doi:10.5281/zenodo.5093701
General format to run the script is
singularity run --cleanenv \
--bind <path_to_noddi_input_folder>:/In_fold \
<path_to_singulairty_image_folder>/noddi-singularity.sif \
/In_fold/<NODDI image> \
/In_fold/<brain_mask> \
/In_fold/<NODDI_roi> \
/In_fold/<bval> \
/In_fold/<bvecs> \
/In_fold/<FittedParams.mat> \
/In_fold/<output_map>
Example
To run the NODDI example dataset
Download the dataset from the above link
singularity run --cleanenv \
--bind /<path_to_noddi_folder>/NODDI_example_dataset:/In_fold \
/<path_to_singulairty_image_folder>/noddi-singularity.sif \
/In_fold/NODDI_DWI.hdr \
/In_fold/roi_mask.hdr \
/In_fold/NODDI_roi.mat\
/In_fold/NODDI_protocol.bval \
/In_fold/NODDI_protocol.bvec \
/In_fold/FittedParams.mat \
/In_fold/example