Nypipe interface for the Lesion-toads approach (Topology-preserving approach to the segmentation of brain images with multiple sclerosis lesions) proposed by Shiee et al. 2010.
-
This interface is based on Nipype. Follow the instructions of the authors webpage.
-
Lesion-toads has to be installed following the provided manual, which is run from the Mipav environment.
-
Then, be sure that MIPAV and the corresponding libraries are in the path. Tipically, MIPAV is installed in the home folder.
JAVALIB=/home/ThisUser/mipav/jre/Contents/Home/lib/ext/
# location of the MIPAV installation to use
MIPAV=/home/ThisUser/mipav
# location of the plugin installation to use
# please replace 'ThisUser' by your user name
PLUGINS=/home/ThisUser/mipav/plugins
So far, the implementation is simple. Just run the method using the provided function. Only the T1 and FLAIR images are available as options. The rest of parameters are run with default values.
from lesion_toads import *
import os
# database dependent
IMAGE_FOLDER='/path/to/your/images'
dirs = os.listdir(IMAGE_FOLDER)
T1='t1name'
FLAIR='flairname'
for f in dirs:
print('processing scan: '+f)
current_t1 = os.path.join(IMAGE_FOLDER,f,T1)
current_flair = os.path.join(IMAGE_FOLDER,f,FLAIR)
lesion_toads(current_t1, current_flair)New parameters can be tuned. Just run the get_help() function to obtain the number of possible parameter to tune.
Wraps command **java edu.jhu.ece.iacl.jist.cli.run edu.jhu.ece.iacl.plugins.classification.MedicAlgorithmLesionToads **
title: Lesion TOADS
category: Developer Tools
description: Algorithm for simulataneous brain structures and MS lesion segmentation of MS Brains. The brain segmentation is topologically consistent and the algorithm can use multiple MR sequences as input data.
N. Shiee, P.-L. Bazin, A.Z. Ozturk, P.A. Calabresi, D.S. Reich, D.L. Pham, "A Topology-Preserving Approach to the Segmentation of Brain Images with Multiple Sclerosis", NeuroImage, vol. 49, no. 2, pp. 1524-1535, 2010.
version: 1.9.R
contributor: Navid Shiee (navid.shiee@nih.gov) http://iacl.ece.jhu.edu/~nshiee/
Inputs::
[Mandatory]
[Optional]
args: (a string)
Additional parameters to the command
flag: %s
environ: (a dictionary with keys which are a value of type 'str' and
with values which are a value of type 'str', nipype default value:
{})
Environment variables
ignore_exception: (a boolean, nipype default value: False)
Print an error message instead of throwing an exception in case the
interface fails to run
inAtlas: ('With Lesion' or 'No Lesion')
Atlas to Use
flag: --inAtlas %s
inAtlas2: (an existing file name)
Atlas File - With Lesions
flag: --inAtlas2 %s
inAtlas3: (an existing file name)
Atlas File - No Lesion - T1 and FLAIR
flag: --inAtlas3 %s
inAtlas4: (an existing file name)
Atlas File - No Lesion - T1 Only
flag: --inAtlas4 %s
inAtlas5: (a float)
Controls the effect of the statistical atlas on the segmentation
flag: --inAtlas5 %f
inAtlas6: ('rigid' or 'multi_fully_affine')
Atlas alignment
flag: --inAtlas6 %s
inConnectivity: ('(26,6)' or '(6,26)' or '(6,18)' or '(18,6)')
Connectivity (foreground,background)
flag: --inConnectivity %s
inCorrect: ('true' or 'false')
Correct MR field inhomogeneity.
flag: --inCorrect %s
inFLAIR: (an existing file name)
FLAIR Image
flag: --inFLAIR %s
inInclude: ('true' or 'false')
Include lesion in WM class in hard classification
flag: --inInclude %s
inMaximum: (an integer (int or long))
Maximum distance from the interventricular WM boundary to downweight
the lesion membership to avoid false postives
flag: --inMaximum %d
inMaximum2: (an integer (int or long))
Maximum Ventircle Distance
flag: --inMaximum2 %d
inMaximum3: (an integer (int or long))
Maximum InterVentricular Distance
flag: --inMaximum3 %d
inMaximum4: (a float)
Maximum amount of relative change in the energy function considered
as the convergence criteria
flag: --inMaximum4 %f
inMaximum5: (an integer (int or long))
Maximum iterations
flag: --inMaximum5 %d
inOutput: ('hard segmentation' or 'hard segmentation+memberships' or
'cruise inputs' or 'dura removal inputs')
Output images
flag: --inOutput %s
inOutput2: ('true' or 'false')
Output the hard classification using maximum membership (not
neceesarily topologically correct)
flag: --inOutput2 %s
inOutput3: ('true' or 'false')
Output the estimated inhomogeneity field
flag: --inOutput3 %s
inSmooting: (a float)
Controls the effect of neighberhood voxels on the membership
flag: --inSmooting %f
inT1_MPRAGE: (an existing file name)
T1_MPRAGE Image
flag: --inT1_MPRAGE %s
inT1_SPGR: (an existing file name)
T1_SPGR Image
flag: --inT1_SPGR %s
null: (a string)
Execution Time
flag: --null %s
outCortical: (a boolean or a file name)
Cortical GM Membership
flag: --outCortical %s
outFilled: (a boolean or a file name)
Filled WM Membership
flag: --outFilled %s
outHard: (a boolean or a file name)
Hard segmentation
flag: --outHard %s
outHard2: (a boolean or a file name)
Hard segmentationfrom memberships
flag: --outHard2 %s
outInhomogeneity: (a boolean or a file name)
Inhomogeneity Field
flag: --outInhomogeneity %s
outLesion: (a boolean or a file name)
Lesion Segmentation
flag: --outLesion %s
outMembership: (a boolean or a file name)
Membership Functions
flag: --outMembership %s
outSulcal: (a boolean or a file name)
Sulcal CSF Membership
flag: --outSulcal %s
outWM: (a boolean or a file name)
WM Mask
flag: --outWM %s
terminal_output: ('stream' or 'allatonce' or 'file' or 'none')
Control terminal output: `stream` - displays to terminal immediately
(default), `allatonce` - waits till command is finished to display
output, `file` - writes output to file, `none` - output is ignored
xDefaultMem: (an integer (int or long))
Set default maximum heap size
flag: -xDefaultMem %d
xMaxProcess: (an integer (int or long), nipype default value: 1)
Set default maximum number of processes.
flag: -xMaxProcess %d
xPrefExt: ('nrrd')
Output File Type
flag: --xPrefExt %s
Outputs::
outCortical: (an existing file name)
Cortical GM Membership
outFilled: (an existing file name)
Filled WM Membership
outHard: (an existing file name)
Hard segmentation
outHard2: (an existing file name)
Hard segmentationfrom memberships
outInhomogeneity: (an existing file name)
Inhomogeneity Field
outLesion: (an existing file name)
Lesion Segmentation
outMembership: (an existing file name)
Membership Functions
outSulcal: (an existing file name)
Sulcal CSF Membership
outWM: (an existing file name)
WM Mask