Diffusion Analyse des Faisceaux

This process compute several quantitative information for each input fascicles bundle.

Description

This process computes scalar features for each input fascicles bundle. The term scalar feature is used as a generic way to refer to a set of scalar values related to one property of an object. For example, a bundle object contains fascicles. Each fascicle has a length. Therefore a length feature can be defined for the bundle. The different values contained in this feature are all related to the length of the fascicles contained in the bundle. This process compute five values for the length feature:

• min: Length of the shortest fascicle.
• max: Length of the longest fascicle.
• mean: Mean length of the fascicles.
• stddev: Standard deviation fo the fascicles length.
• median: Median length of the fascicles.

Two kind of features on each bundle of fascicles:

• Morphological features: To date only fascicle length is used but the morphological feature set is likely to enlarge in the future. The length feature contains the values described above.
• Image projection features: Each fascicle is projected in the various images given in parameter of this process. Then statistics about each image value along the projected fascicle are extracted. Each image given in parameter produce one feature:
  • adc: Apparent diffusion coefficient.
  • fa: Fractional anisotropy.
  • vr: Volume ratio.
  • stddev: Diffusion standard deviation.
  • vr: Diffusion weighted T2.
  Each image related feature contains statistics computed on the image (min, max, mean, stddev and median). These values are computed for all the points of all the fascicles of each bundle (as if each bundle was used as a ROI).
  If the process parameter slices is used, statistics along each bundle length are produced. Each bundle is cut in slices slices. Then, each slice point is projected in the images and the same values as bundle statistics are computed for each slice (min, max, mean, stddev and median). In this case, the feature contains both one bundle global value and one vector of slices values for each statistic.

The result of this process is stored in a feature_1.0 format (which is a minf format and can therefore use either a Python syntax or a XML syntax). This result is organised as a hierarchy. The first level of this hierarchy contains one element per bundle. On the second level (i.e. for each bundle), there is one element for each feature. On the third level ( i.e. for each feature), there is one element (containing a single number) for each statistic function (min, max, mean, stddev and median) and one optional element called _vectors (if slices is used) containing one vector of numbers for each statistic function (and an abscissa element which can be used to display these vectors as graphs). Here is an example of a feature_1.0 file computed on a bundle set containing 2 bundles (roi_1 and roi_2) and with slices = 10:

---

attributes = {
  'format': 'features_1.0',
  'content_type': 'bundles_features',
  'roi_01': {
    'adc': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 8.75151e-10, 7.8922e-10, 7.9116e-10, 7.63253e-10, 7.51277e-10, 7.7328e-10, 7.71915e-10, 7.72216e-10, 7.88993e-10, 8.01587e-10, ],
        'stddev': [ 2.6872e-10, 5.54337e-11, 5.35846e-11, 3.8797e-11, 5.03774e-11, 3.9858e-11, 4.18283e-11, 4.06521e-11, 5.86814e-11, 3.09343e-10, ],
        'min': [ 5.1147e-10, 6.85862e-10, 6.82759e-10, 7.03226e-10, 6.35999e-10, 6.66687e-10, 6.99779e-10, 7.10966e-10, 6.90667e-10, 3.52562e-11, ],
        'max': [ 1.90512e-09, 9.2389e-10, 9.15974e-10, 8.46003e-10, 8.79147e-10, 8.58503e-10, 8.73618e-10, 8.8356e-10, 9.28362e-10, 1.7861e-09, ],
        'median': [ 8.0669e-10, 7.85439e-10, 7.81571e-10, 7.60641e-10, 7.50846e-10, 7.71922e-10, 7.6772e-10, 7.67205e-10, 7.91226e-10, 8.14963e-10, ],
      },
      'mean': 7.86453e-10,
      'stddev': 1.0537e-10,
      'min': 3.52562e-11,
      'max': 2.00021e-09,
      'median': 7.7649e-10,
    },
    'dwt2': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 443.892, 422.896, 416.619, 417.83, 414.836, 413.76, 414.646, 413.964, 419.753, 351.112, ],
        'stddev': [ 70.244, 39.1324, 27.4811, 23.1174, 21.4667, 25.2494, 27.2151, 22.0603, 17.74, 112.446, ],
        'min': [ 165.674, 357.585, 315.187, 356.166, 363.467, 340.483, 341.984, 329.286, 356.265, 28.465, ],
        'max': [ 571.684, 535.659, 453.59, 451.237, 440.847, 448.838, 443.863, 447.302, 464.549, 440.419, ],
        'median': [ 444.106, 418.882, 420.224, 420.142, 421.6, 418.705, 425.239, 419.093, 418.105, 407.313, ],
      },
      'mean': 420.76,
      'stddev': 36.9771,
      'min': 28.465,
      'max': 603.872,
      'median': 424.657,
    },
    'fa': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 0.511815, 0.496629, 0.604314, 0.628366, 0.59314, 0.523773, 0.494426, 0.516998, 0.423428, 0.257682, ],
        'stddev': [ 0.135839, 0.154436, 0.127339, 0.0921732, 0.126183, 0.12906, 0.083203, 0.0843948, 0.101978, 0.156139, ],
        'min': [ 0.214012, 0.243474, 0.305914, 0.375954, 0.252565, 0.29064, 0.349038, 0.327365, 0.226794, 0.00167768, ],
        'max': [ 0.77341, 0.752725, 0.772975, 0.771586, 0.790421, 0.73756, 0.643487, 0.65627, 0.603582, 0.481977, ],
        'median': [ 0.488002, 0.480083, 0.627162, 0.642, 0.607358, 0.524098, 0.488491, 0.53764, 0.404398, 0.285002, ],
      },
      'mean': 0.514956,
      'stddev': 0.139625,
      'min': 0.00167768,
      'max': 0.83117,
      'median': 0.525463,
    },
    'stddev': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 2.9076e-10, 2.14055e-10, 3.0577e-10, 3.50443e-10, 3.36184e-10, 2.90465e-10, 2.61997e-10, 2.7067e-10, 2.14319e-10, 1.37787e-10, ],
        'stddev': [ 1.35863e-10, 1.15343e-10, 1.25589e-10, 9.10421e-11, 9.04655e-11, 1.0705e-10, 7.72566e-11, 8.04067e-11, 5.83886e-11, 9.22118e-11, ],
        'min': [ 1.16895e-10, 5.24838e-11, 5.3487e-11, 1.17174e-10, 1.53199e-10, 1.05018e-10, 1.26519e-10, 9.88504e-11, 6.82857e-11, 2.81754e-12, ],
        'max': [ 5.76015e-10, 4.76216e-10, 5.36406e-10, 5.19542e-10, 5.12446e-10, 5.10919e-10, 4.18639e-10, 4.19344e-10, 3.20628e-10, 3.26826e-10, ],
        'median': [ 2.46676e-10, 2.02138e-10, 3.11621e-10, 3.57696e-10, 3.18832e-10, 2.90912e-10, 2.49506e-10, 2.85428e-10, 2.2151e-10, 1.26879e-10, ],
      },
      'mean': 2.67689e-10,
      'stddev': 1.06195e-10,
      'min': 1.81471e-12,
      'max': 5.91988e-10,
      'median': 2.69958e-10,
    },
    'vr': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 0.299651, 0.289705, 0.42047, 0.436608, 0.40042, 0.328453, 0.270725, 0.300002, 0.211442, 0.101387, ],
        'stddev': [ 0.149027, 0.167209, 0.159751, 0.12273, 0.153615, 0.171637, 0.0946103, 0.0930927, 0.0967568, 0.0864733, ],
        'min': [ 0.0557933, 0.059842, 0.101457, 0.159361, 0.0681085, 0.0883092, 0.123964, 0.116549, 0.0579234, 0.000132187, ],
        'max': [ 0.641994, 0.610262, 0.647237, 0.630444, 0.723072, 0.760691, 0.442624, 0.47167, 0.416598, 0.293256, ],
        'median': [ 0.258129, 0.236291, 0.443339, 0.434753, 0.399115, 0.307833, 0.25958, 0.303568, 0.181795, 0.0945587, ],
      },
      'mean': 0.30949,
      'stddev': 0.152233,
      'min': 0.000132187,
      'max': 0.791289,
      'median': 0.297167,
    },
    'length': {
      'mean': 69.4554,
      'stddev': 24.695,
      'min': 28.5937,
      'max': 112.5,
      'median': 77.8125,
    },
  },
  'roi_02': {
    'adc': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 7.55662e-10, 7.96565e-10, 7.82272e-10, 7.77354e-10, 7.73943e-10, 7.69115e-10, 7.7612e-10, 8.14012e-10, 8.10743e-10, 9.58386e-10, ],
        'stddev': [ 4.50403e-10, 4.26217e-11, 5.45371e-11, 4.33616e-11, 5.14154e-11, 4.09276e-11, 4.64212e-11, 1.43277e-10, 1.38865e-10, 5.77474e-10, ],
        'min': [ 9.59585e-13, 7.35933e-10, 6.89633e-10, 6.87511e-10, 6.78084e-10, 6.53927e-10, 6.95704e-10, 7.2073e-10, 6.4995e-10, 6.18574e-11, ],
        'max': [ 2.50078e-09, 8.7775e-10, 9.10358e-10, 8.59984e-10, 8.7873e-10, 8.33277e-10, 8.89597e-10, 1.57785e-09, 1.42435e-09, 3.72835e-09, ],
        'median': [ 7.86301e-10, 8.02156e-10, 7.81642e-10, 7.79362e-10, 7.70891e-10, 7.71519e-10, 7.75616e-10, 7.9634e-10, 7.93467e-10, 8.25472e-10, ],
      },
      'mean': 7.93734e-10,
      'stddev': 1.09571e-10,
      'min': 9.59585e-13,
      'max': 3.72835e-09,
      'median': 7.82111e-10,
    },
    'dwt2': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 333.191, 418.477, 416.837, 417.536, 414.326, 412.153, 411.785, 405.289, 411.277, 404.563, ],
        'stddev': [ 141.479, 16.8872, 22.9798, 20.479, 26.5082, 21.8279, 20.5653, 41.0517, 40.3369, 102.338, ],
        'min': [ 39.0871, 385.93, 319.668, 352.065, 325.444, 326.714, 365.613, 219.101, 279.206, 49.4523, ],
        'max': [ 446.521, 468.025, 439.407, 440.123, 441.916, 442.182, 458.386, 450.591, 501.788, 538.31, ],
        'median': [ 410.499, 420.74, 424.093, 424.597, 426.4, 419.474, 420.086, 413.785, 412.526, 431.846, ],
      },
      'mean': 416.292,
      'stddev': 32.6972,
      'min': 39.0871,
      'max': 538.31,
      'median': 421.23,
    },
    'fa': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 0.25885, 0.440121, 0.47062, 0.481671, 0.513904, 0.57113, 0.621281, 0.604215, 0.522728, 0.442348, ],
        'stddev': [ 0.162208, 0.12572, 0.103203, 0.124597, 0.125426, 0.114154, 0.112637, 0.13612, 0.152782, 0.171637, ],
        'min': [ 2.48595e-05, 0.248728, 0.265612, 0.286785, 0.288229, 0.332387, 0.393642, 0.314669, 0.233174, 0.0107373, ],
        'max': [ 0.549464, 0.776157, 0.659183, 0.704337, 0.717392, 0.750768, 0.810673, 0.806267, 0.763482, 0.714531, ],
        'median': [ 0.304338, 0.454821, 0.477216, 0.511547, 0.573017, 0.608326, 0.650218, 0.631319, 0.534586, 0.455771, ],
      },
      'mean': 0.514779,
      'stddev': 0.143334,
      'min': 2.48595e-05,
      'max': 0.825248,
      'median': 0.514497,
    },
    'stddev': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 1.19706e-10, 2.45966e-10, 2.35657e-10, 2.44345e-10, 2.50864e-10, 2.96496e-10, 3.2689e-10, 3.08604e-10, 2.33414e-10, 2.0985e-10, ],
        'stddev': [ 7.8038e-11, 9.74333e-11, 7.45896e-11, 9.35785e-11, 1.0075e-10, 8.84268e-11, 9.49522e-11, 1.11094e-10, 9.50538e-11, 1.12851e-10, ],
        'min': [ 3.38967e-14, 1.08523e-10, 7.91196e-11, 7.33858e-11, 6.10919e-11, 1.09162e-10, 1.43504e-10, 1.05246e-10, 7.93635e-11, 6.01953e-12, ],
        'max': [ 2.60609e-10, 5.34382e-10, 3.71282e-10, 4.21096e-10, 4.57033e-10, 4.81864e-10, 5.05082e-10, 5.30173e-10, 4.36367e-10, 4.9898e-10, ],
        'median': [ 1.31989e-10, 2.61754e-10, 2.42623e-10, 2.29916e-10, 2.46681e-10, 3.05931e-10, 3.54037e-10, 3.48675e-10, 2.4594e-10, 1.87596e-10, ],
      },
      'mean': 2.55931e-10,
      'stddev': 1.0251e-10,
      'min': 3.38967e-14,
      'max': 6.22208e-10,
      'median': 2.49369e-10,
    },
    'vr': {
      '_vectors': {
        'abscissa': [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, ],
        'mean': [ 0.107356, 0.222856, 0.250605, 0.26674, 0.301366, 0.365061, 0.431092, 0.410786, 0.318461, 0.245799, ],
        'stddev': [ 0.0916894, 0.123492, 0.11069, 0.141392, 0.137019, 0.134859, 0.149091, 0.167859, 0.167866, 0.149612, ],
        'min': [ 1.2666e-06, 0.063185, 0.0772016, 0.0789892, 0.0909701, 0.121205, 0.163633, 0.0984095, 0.0576192, 0.00185344, ],
        'max': [ 0.305612, 0.638708, 0.487381, 0.580751, 0.553781, 0.605434, 0.73563, 0.717495, 0.650704, 0.519696, ],
        'median': [ 0.0972249, 0.231472, 0.240343, 0.26659, 0.33249, 0.392522, 0.443441, 0.427296, 0.309672, 0.229115, ],
      },
      'mean': 0.306806,
      'stddev': 0.156608,
      'min': 1.2666e-06,
      'max': 0.751079,
      'median': 0.283922,
    },
    'length': {
      'mean': 69.4482,
      'stddev': 22.3767,
      'min': 29.5313,
      'max': 107.344,
      'median': 73.125,
    },
  },
}

---

Features files can be viewed by clicking on the eye icon which calls Show Scalar Features.

Paramètres

bundles: Fascicles bundles ( entrée )

fascicles bundle set file

bundles_transformation: Transformation matrix ( optional, entrée )

Referential transformation applied to each fascicle point.

minimum_length: Réel ( optional, input )

Remove all fascicles shorter than minimum_length millimeters.

slices: Entier ( optional, input )

Each fascicles bundle is cut in slices parts to get information along the bundle trajectory.

apparent_diffusion_coefficient: Apparent Diffusion Coefficient ( optional, entrée )

If this parameter is given, apparent diffusion coefficient values along the fascicles are included in output features

fractional_anisotropy: Fractional Anisotropy ( optional, entrée )

If this parameter is given, fractional anisotropy values along the fascicles are included in output features

volume_ratio: Volume Ratio ( optional, entrée )

If this parameter is given, volume ratio values along the fascicles are included in output features

standard_deviation: Diffusion Standard Deviation ( optional, entrée )

If this parameter is given, standard deviation image values along the fascicles are included in output features

diffusion_weighted_t2: Diffusion Weighted T2 ( optional, entrée )

If this parameter is given, diffusion weighted T2 values along the fascicles are included in output features

statistics: Bundles scalar features ( sortie )

Output file containing extracted features (i.e. numerical values)

Informations techniques

Toolbox : Diffusion et Tractographie

Niveau d'utilisateur : 1

Identifiant : DiffusionBundleAnalysis

Nom de fichier : brainvisa/toolboxes/connectomist/processes/Tracking Pipeline Components/DiffusionBundleAnalysis.py

Supported file formats :

bundles :

Aims bundles

bundles_transformation :

Transformation matrix

apparent_diffusion_coefficient :

GIS image, VIDA image, NIFTI-1 image, MINC image, gz compressed MINC image, DICOM image, TIFF image, XBM image, PBM image, PGM image, BMP image, XPM image, PPM image, gz compressed NIFTI-1 image, TIFF(.tif) image, ECAT i image, PNG image, JPEG image, MNG image, GIF image, SPM image, ECAT v image

fractional_anisotropy :

GIS image, VIDA image, NIFTI-1 image, MINC image, gz compressed MINC image, DICOM image, TIFF image, XBM image, PBM image, PGM image, BMP image, XPM image, PPM image, gz compressed NIFTI-1 image, TIFF(.tif) image, ECAT i image, PNG image, JPEG image, MNG image, GIF image, SPM image, ECAT v image

volume_ratio :

GIS image, VIDA image, NIFTI-1 image, MINC image, gz compressed MINC image, DICOM image, TIFF image, XBM image, PBM image, PGM image, BMP image, XPM image, PPM image, gz compressed NIFTI-1 image, TIFF(.tif) image, ECAT i image, PNG image, JPEG image, MNG image, GIF image, SPM image, ECAT v image

standard_deviation :

GIS image, VIDA image, NIFTI-1 image, MINC image, gz compressed MINC image, DICOM image, TIFF image, XBM image, PBM image, PGM image, BMP image, XPM image, PPM image, gz compressed NIFTI-1 image, TIFF(.tif) image, ECAT i image, PNG image, JPEG image, MNG image, GIF image, SPM image, ECAT v image

diffusion_weighted_t2 :

GIS image, VIDA image, NIFTI-1 image, MINC image, gz compressed MINC image, DICOM image, TIFF image, XBM image, PBM image, PGM image, BMP image, XPM image, PPM image, gz compressed NIFTI-1 image, TIFF(.tif) image, ECAT i image, PNG image, JPEG image, MNG image, GIF image, SPM image, ECAT v image

statistics :

Aims scalar features