Merge pull request #3 from chrisfilo/fix/mrtrix

Fix/mrtrix

Author: swederik

examples/mrtrix_dti_tutorial.py

@@ -23,15 +23,9 @@
 import nipype.pipeline.engine as pe          # pypeline engine
 import nipype.interfaces.mrtrix as mrtrix   #<---- The important new part!
 import nipype.interfaces.fsl as fsl
-import nibabel as nb
+import nipype.algorithms.misc as misc
 import os, os.path as op                     # system functions
 
-"""
-We import the voxel-, data-, and affine-grabbing functions from the Camino DTI processing workflow
-"""
-
-from nipype.workflows.camino.diffusion import get_vox_dims, get_data_dims, get_affine
-
 """
 This needs to point to the fdt folder you can find after extracting
 	
@@ -50,7 +44,7 @@
 infosource = pe.Node(interface=util.IdentityInterface(fields=['subject_id']), name="infosource")
 infosource.iterables = ('subject_id', subject_list)
 
-info = dict(dwi=[['subject_id', 'dwi']],
+info = dict(dwi=[['subject_id', 'data']],
             bvecs=[['subject_id','bvecs']],
             bvals=[['subject_id','bvals']])
 
@@ -65,7 +59,7 @@
 
 datasource.inputs.template = "%s/%s"
 datasource.inputs.base_directory = data_dir
-datasource.inputs.field_template = dict(dwi='%s/%s.nii')
+datasource.inputs.field_template = dict(dwi='%s/%s.nii.gz')
 datasource.inputs.template_args = info
 
 """
@@ -99,7 +93,7 @@
 	* Apparent diffusion coefficient
 	* Fractional anisotropy
 """
-
+gunzip = pe.Node(interface=misc.Gunzip(), name='gunzip')
 dwi2tensor = pe.Node(interface=mrtrix.DWI2Tensor(),name='dwi2tensor')
 tensor2vector = pe.Node(interface=mrtrix.Tensor2Vector(),name='tensor2vector')
 tensor2adc = pe.Node(interface=mrtrix.Tensor2ApparentDiffusion(),name='tensor2adc')
@@ -179,7 +173,8 @@
 
 tractography.connect([(inputnode, fsl2mrtrix, [("bvecs", "bvec_file"),
 												("bvals", "bval_file")])])
-tractography.connect([(inputnode, dwi2tensor,[("dwi","in_file")])])
+tractography.connect([(inputnode, gunzip,[("dwi","in_file")])])
+tractography.connect([(gunzip, dwi2tensor,[("out_file","in_file")])])
 tractography.connect([(fsl2mrtrix, dwi2tensor,[("encoding_file","encoding_file")])])
 
 tractography.connect([(dwi2tensor, tensor2vector,[['tensor','in_file']]),
@@ -193,7 +188,7 @@
 fractional anisotropy image, and thresholds it to get the single-fiber voxels.
 """
 
-tractography.connect([(inputnode, MRconvert,[("dwi","in_file")])])
+tractography.connect([(gunzip, MRconvert,[("out_file","in_file")])])
 tractography.connect([(MRconvert, threshold_b0,[("converted","in_file")])])
 tractography.connect([(threshold_b0, median3d,[("out_file","in_file")])])
 tractography.connect([(median3d, erode_mask_firstpass,[("out_file","in_file")])])
@@ -206,8 +201,8 @@
 Here the thresholded white matter mask is created for seeding the tractography.
 """
 
-tractography.connect([(inputnode, bet,[("dwi","in_file")])])
-tractography.connect([(inputnode, gen_WM_mask,[("dwi","in_file")])])
+tractography.connect([(gunzip, bet,[("out_file","in_file")])])
+tractography.connect([(gunzip, gen_WM_mask,[("out_file","in_file")])])
 tractography.connect([(bet, gen_WM_mask,[("mask_file","binary_mask")])])
 tractography.connect([(fsl2mrtrix, gen_WM_mask,[("encoding_file","encoding_file")])])
 tractography.connect([(gen_WM_mask, threshold_wmmask,[("WMprobabilitymap","in_file")])])
@@ -216,15 +211,15 @@
 Next we estimate the fiber response distribution.
 """
 
-tractography.connect([(inputnode, estimateresponse,[("dwi","in_file")])])
+tractography.connect([(gunzip, estimateresponse,[("out_file","in_file")])])
 tractography.connect([(fsl2mrtrix, estimateresponse,[("encoding_file","encoding_file")])])
 tractography.connect([(threshold_FA, estimateresponse,[("out_file","mask_image")])])
 
 """
 Run constrained spherical deconvolution.
 """
 
-tractography.connect([(inputnode, csdeconv,[("dwi","in_file")])])
+tractography.connect([(gunzip, csdeconv,[("out_file","in_file")])])
 tractography.connect([(gen_WM_mask, csdeconv,[("WMprobabilitymap","mask_image")])])
 tractography.connect([(estimateresponse, csdeconv,[("response","response_file")])])
 tractography.connect([(fsl2mrtrix, csdeconv,[("encoding_file","encoding_file")])])
@@ -236,7 +231,10 @@
 tractography.connect([(threshold_wmmask, probCSDstreamtrack,[("out_file","seed_file")])])
 tractography.connect([(csdeconv, probCSDstreamtrack,[("spherical_harmonics_image","in_file")])])
 tractography.connect([(probCSDstreamtrack, tracks2prob,[("tracked","in_file")])])
-tractography.connect([(inputnode, tracks2prob,[("dwi","template_file")])])
+tractography.connect([(gunzip, tracks2prob,[("out_file","template_file")])])
+
+tractography.connect([(gunzip, tck2trk,[("out_file","image_file")])])
+tractography.connect([(probCSDstreamtrack, tck2trk,[("tracked","in_file")])])
 
 """
 Finally, we create another higher-level workflow to connect our tractography workflow with the info and datagrabbing nodes

nipype/algorithms/misc.py

@@ -456,3 +456,36 @@ def _list_outputs(self):
         if isdefined(self.inputs.regress_poly):
             outputs['detrended_file'] = self._gen_output_file_name('detrended')
         return outputs
+
+class GunzipInputSpec(BaseInterfaceInputSpec):
+    in_file = File(exists=True, mandatory=True)
+
+class GunzipOutputSpec(TraitedSpec):
+    out_file = File(exists=True)
+
+class Gunzip(BaseInterface):
+    """
+
+    """
+    input_spec = GunzipInputSpec
+    output_spec = GunzipOutputSpec
+
+    def _gen_output_file_name(self):
+        _, base, ext = split_filename(self.inputs.in_file)
+        if ext[-2:].lower() == ".gz":
+            ext = ext[:-3]
+        return os.path.abspath(base + ext[:-3])
+
+    def _run_interface(self, runtime):
+        import gzip
+        in_file = gzip.open(self.inputs.in_file, 'rb')
+        out_file = open(self._gen_output_file_name(), 'wb')
+        out_file.write(in_file.read())
+        out_file.close()
+        in_file.close()
+        return runtime
+
+    def _list_outputs(self):
+        outputs = self._outputs().get()
+        outputs['out_file'] = self._gen_output_file_name()
+        return outputs

nipype/interfaces/mrtrix/convert.py

@@ -328,7 +328,6 @@ def _run_interface(self, runtime):
         # as a commandline argument to the matlab executable (without creating any files).
         # This, however, is less reliable and harder to debug (code will be reduced to
         # a single line and stripped of any comments).
-        print script
         result = MatlabCommand(script=script, mfile=True)
         r = result.run()
         return runtime