WMTI-watson-py

Given mean, axial, radial diffusivity, mean, axial, radial kurtosis maps, calculate WM model parameter maps:

• f: axonal water Fraction.
• Da: Axonal Diffusivity.
• Depar: Extra-axonal PARallel Diffusivity.
• Deperp: Extra-axonal PERPendicular Diffusivity.
• c2: mean cos ^ 2 of the axon orientation dispersion [1/3, 1].
  • c2 = 1 / 3 fully isotropic,
  • c2 = 1 perfectly parallel.

c2 is directly related to the Watson distribution concentration parameter kappa(same as in NODDI).

How To:

WMTI-watson-py requires a path to a folder containing the mean, axial and radial diffusivity as well as the mean, axial and radial kurtosis maps, all of which must be identifiable from the acronyms md, ad, rd, mk, ak, rk, respectively. Alternatively, an ordered tuple or list containing the paths to each image can be fed, must follow the order: ('md', 'ad', 'rd', 'mk', 'ak', 'rk').

import os
from WMTI_Watson import WMTI_Watson

# Prepare inputs

input_path = '/home/data/dataset1/subject1/dkifit'
output_path = '/home/data/dataset1/subject1/wmtiwatson'
mask = '/home/data/dataset1/subject1/mask.nii.gz'
os.mkdir(output_path)
print(os.listdir(path))

Out[1]: ['md.nii.gz', 'ad.nii.gz', 'rd.nii.gz', 'mk.nii.gz', 'ak.nii.gz', 'rk.nii.gz']

# Initialize the wrapper class
wmti = WMTI_Watson(path, mask=mask, params='invivo', nodes=4)
# Fit
wmti.fit()

Out[2]: 'Completed in 334s'

# Retreive fitted maps as numpy arrays..
f, Da, Depar, Deperp, c2 = wmti.maps()
# ..or save the maps
wmti.save(output_path)
print(os.listdir(output_path))

Out[3]: ['Depar.nii.gz', 'Deperp.nii.gz', 'f.nii.gz', 'c2.nii.gz', 'Da.nii.gz']

Internal functions

WMTI-watson-py internal function are easily accessible via:

import WMTI_Watson as wmti
wmti.wmti_watson_f
wmti.WMTI_Watson_maps
wmti.parfit_wmti_watson
wmti.normal_fit_wmti_watson