diffusion_FA

Our deep learning network allows to map Fractional Anisotropy form a set of only 10 Diffusion Weighed volumes.

Give your 10 diffusion weighed images and obtein the FA!!

For all details check the paper "A generalized deep learning network for fractional anisotropy reconstruction : Application to epilepsy and multiple sclerosis" (Gaviraghi et al, 2022) https://doi.org/10.3389/fninf.2022.891234 If you use this code cite this article!

you need

• Python3 (you need the following packages: natsort, tensorflow, keras, scipy, numpy)

• Matlab

For the compatibility between the Python and Matlab version see https://www.mathworks.com/content/dam/mathworks/mathworks-dot-com/support/sysreq/files/python-compatibility.pdf

Open a terminal and navigate in the foldere where you want work ('path_to_down').

Clone the code in your 'path_to_down'

git clone [email protected]:marta-gaviraghi/diffusion_FA.git

Give your 10 diffusion weighet images and wait ...

python3 main_python.py 'path_dir' 'path_to_down' 'dwi_10'

if you have the brain mask you can give it as input

python3 main_python.py 'path_dir' 'path_to_down' 'dwi_10' 'brain_mask'

where the input are:

MANDATORY INPUTS:

• path_dir = 'path to the folder that contains the directory of each subject or a dirctory of subject'
• path_to_down = 'path in where you clone the repository'
• dwi_10 = 'name of the nifti of the 10 DW volumes'

OPTIONAL

• brain_mask = 'name of the brain mask'

Remember to define your path and name of fie between the single comma and that for your name file you don't indicate the extention .nii.gz