qDLab

Quantitative Diffusion MRI Made Easy with qDLab: a Software for Data Visualization and Analysis of 1D qspace sampling

author : Tanguy Duval
contact : tanguy.duval@polymtl.ca
institution : NeuroPoly Lab, Polytechnique Montreal date : 2016

Installation

download and open the app (qDLab.mlappinstall) : https://github.com/neuropoly/qDLab/releases/
Launch the app from Matlab App tab
The app will propose you to open a NIFTI file. Open one of the two example dataset.
The app will propose you to open a schemefile. Open one of the two example dataset.
select a pixel and press enter --> the MRI signal in this particular pixel will be plotted
Click on "model fitting" button --> CHARMED model will be fitted on the data. Results are displayed in the Matlab console
Click on "Generate Map" button --> A folder will be generated with all the fitted maps.

Enable NODDI

install the NODDI toolbox and add it to your Matlab path: http://www.nitrc.org/projects/noddi_toolbox

Support

Use GitHub issue to report any error or for help. Expect an answer in <24h.

Schemefile

The schemefile describes the diffusion protocol associated to your dataset. Example of schemefile : 336810.scheme

You can generate your schemefile using the function scd_schemefile_create.
See Examples:
>> help scd_schemefile_create
If acquired two datasets with the same bvecs.txt but different diffusion times (20 and 35ms):

bvecs_files = {'bvecs.txt', 'bvecs.txt'};   
add_b0_beginning = 0;   
DELTA       = [20 35]*1e-3; % s  
delta       = [8  8 ]*1e-3; % s   
Gmax        = [380 380]*1e-3; % T/m   
TE          = [69 70]*1e-3; %s   
acq_basename= date;   
scd_schemefile_create(bvecs_files, add_b0_beginning, DELTA, delta,Gmax, TE, acq_basename);

For more info see the description of the STEJSKALTANNER schemefile on the following link: http://camino.cs.ucl.ac.uk/index.php?n=Docs.SchemeFiles

Add your own model

New quantitative Diffusion models can be easily added to qDLab by adding a Matlab file (.m) in the folder qDLab/code/CUSTOM/
As an example, select CUSTOM_MODEL.m in qDLab.
Note that options can easily be added to the GUI. See qDLab/code/CUSTOM/NODDI.m

Features

Preview data in 1 voxel + fitting

Compare 2 voxels

Subsample your dataset (e.g. Gmax = 40mT/m) and visualize them as you like (e.g. Xaxis : bvalue or qvalue?)

Generate Maps

Preprocessing

Use your favourite software for motion correction, distortion correction, and merge your diffusion data in one NIFTI file (also merge your schemefiles accordingly!)

neuropoly/qDLab