added denoise workflow

Author: oesteban

nipype/workflows/dmri/__init__.py

@@ -1 +1 @@
-import camino, mrtrix, fsl
+import camino, mrtrix, fsl, dipy

nipype/workflows/dmri/dipy/__init__.py

@@ -0,0 +1,5 @@
+# coding: utf-8
+# emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: nil -*-
+# vi: set ft=python sts=4 ts=4 sw=4 et:
+
+from denoise import nlmeans_pipeline

nipype/workflows/dmri/dipy/denoise.py

@@ -0,0 +1,119 @@
+# coding: utf-8
+# emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: nil -*-
+# vi: set ft=python sts=4 ts=4 sw=4 et:
+
+import nipype.pipeline.engine as pe
+import nipype.interfaces.utility as niu
+from nipype.interfaces import fsl
+from nipype.interfaces import dipy
+
+
+def nlmeans_pipeline(name='Denoise',
+                     params={'patch_radius': 1, 'block_radius': 5}):
+    """
+    Workflow that performs nlmeans denoising
+
+    Example
+    -------
+
+    >>> from nipype.workflows.dmri.dipy.denoise import nlmeans_pipeline
+    >>> denoise = nlmeans_pipeline()
+    >>> denoise.inputs.inputnode.in_file = 'diffusion.nii'
+    >>> denoise.inputs.inputnode.in_mask = 'mask.nii'
+    >>> denoise.run() # doctest: +SKIP
+
+
+    """
+    inputnode = pe.Node(niu.IdentityInterface(fields=['in_file', 'in_mask']),
+                        name='inputnode')
+    outputnode = pe.Node(niu.IdentityInterface(fields=['out_file']),
+                         name='outputnode')
+
+    nmask = pe.Node(niu.Function(input_names=['in_file', 'in_mask'],
+                    output_names=['out_file'], function=bg_mask),
+                    name='NoiseMsk')
+    nlmeans = pe.Node(dipy.Denoise(**params), name='NLMeans')
+
+    wf = pe.Workflow(name=name)
+    wf.connect([
+         (inputnode,  nmask,       [('in_file', 'in_file'),
+                                    ('in_mask', 'in_mask')])
+        ,(inputnode,  nlmeans,     [('in_file', 'in_file'),
+                                    ('in_mask', 'in_mask')])
+        ,(nmask,      nlmeans,     [('out_file', 'noise_mask')])
+        ,(nlmeans,    outputnode,  [('out_file', 'out_file')])
+    ])
+    return wf
+
+
+def csf_mask(in_file, in_mask, out_file=None):
+    """
+    Artesanal mask of csf in T2w-like images
+    """
+    import nibabel as nb
+    import numpy as np
+    from scipy.ndimage import binary_erosion, binary_opening, label
+    import scipy.ndimage as nd
+    import os.path as op
+
+    if out_file is None:
+        fname,ext = op.splitext(op.basename(in_file))
+        if ext == ".gz":
+            fname,ext2 = op.splitext(fname)
+            ext = ext2 + ext
+        out_file = op.abspath("%s_csfmask%s" % (fname, ext))
+
+    im = nb.load(in_file)
+    hdr = im.get_header().copy()
+    hdr.set_data_dtype(np.uint8)
+    hdr.set_xyzt_units('mm')
+    imdata = im.get_data()
+    msk = nb.load(in_mask).get_data()
+    msk = binary_erosion(msk,
+                         structure=np.ones((15, 15, 10))).astype(np.uint8)
+    thres = np.percentile(imdata[msk > 0].reshape(-1), 90.0)
+    imdata[imdata < thres] = 0
+    imdata = imdata * msk
+    imdata[imdata > 0] = 1
+    imdata = binary_opening(imdata,
+                            structure=np.ones((2, 2, 2))).astype(np.uint8)
+
+    label_im, nb_labels = label(imdata)
+    sizes = nd.sum(imdata, label_im, range(nb_labels + 1))
+    mask_size = sizes != sizes.max()
+    remove_pixel = mask_size[label_im]
+    label_im[remove_pixel] = 0
+    label_im[label_im > 0] = 1
+    nb.Nifti1Image(label_im.astype(np.uint8),
+                   im.get_affine(), hdr).to_filename(out_file)
+    return out_file
+
+
+def bg_mask(in_file, in_mask, out_file=None):
+    """
+    Rough mask of background from brain masks
+    """
+    import nibabel as nb
+    import numpy as np
+    from scipy.ndimage import binary_dilation
+    import scipy.ndimage as nd
+    import os.path as op
+
+    if out_file is None:
+        fname,ext = op.splitext(op.basename(in_file))
+        if ext == ".gz":
+            fname,ext2 = op.splitext(fname)
+            ext = ext2 + ext
+        out_file = op.abspath("%s_bgmask%s" % (fname, ext))
+
+    im = nb.load(in_file)
+    hdr = im.get_header().copy()
+    hdr.set_data_dtype(np.uint8)
+    hdr.set_xyzt_units('mm')
+    imdata = im.get_data()
+    msk = nb.load(in_mask).get_data()
+    msk = 1 - binary_dilation(msk,
+                              structure=np.ones((20, 20, 20)))
+    nb.Nifti1Image(msk.astype(np.uint8),
+                   im.get_affine(), hdr).to_filename(out_file)
+    return out_file

nipype/workflows/dmri/setup.py

@@ -1,6 +1,8 @@
 # emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: nil -*-
 # vi: set ft=python sts=4 ts=4 sw=4 et:
-def configuration(parent_package='',top_path=None):
+
+
+def configuration(parent_package='', top_path=None):
     from numpy.distutils.misc_util import Configuration
 
     config = Configuration('dmri', parent_package, top_path)
@@ -9,6 +11,7 @@ def configuration(parent_package='',top_path=None):
     config.add_subpackage('mrtrix')
     config.add_subpackage('fsl')
     config.add_subpackage('connectivity')
+    config.add_subpackage('dipy')
 
     return config