Add EM tissues segmentation script

em_segmentation/animaAtlasEMTissuesSegmentation.py

@@ -0,0 +1,174 @@
+#!/usr/bin/python3
+# Warning: works only on unix-like systems, not windows where "python3 animaAtlasEMTissuesSegmentation.py ..." has to be run
+
+import sys
+import argparse
+
+if sys.version_info[0] > 2:
+    import configparser as ConfParser
+else:
+    import ConfigParser as ConfParser
+
+import glob
+import os
+from shutil import copyfile, rmtree
+from subprocess import call, check_output
+import uuid
+
+configFilePath = os.path.join(os.path.expanduser("~"), ".anima",  "config.txt")
+if not os.path.exists(configFilePath):
+    print('Please create a configuration file for Anima python scripts. Refer to the README')
+    quit()
+
+configParser = ConfParser.RawConfigParser()
+configParser.read(configFilePath)
+
+animaDir = configParser.get("anima-scripts", 'anima')
+animaExtraDataDir = configParser.get("anima-scripts", 'extra-data-root')
+animaPyramidalBMRegistration = os.path.join(animaDir, "animaPyramidalBMRegistration")
+animaDenseSVFBMRegistration = os.path.join(animaDir, "animaDenseSVFBMRegistration")
+animaTransformSerieXmlGenerator = os.path.join(animaDir, "animaTransformSerieXmlGenerator")
+animaApplyTransformSerie = os.path.join(animaDir, "animaApplyTransformSerie")
+animaTissuesEMClassification = os.path.join(animaDir, "animaTissuesEMClassification")
+animaConvertImage = os.path.join(animaDir, "animaConvertImage")
+animaMaskImage = os.path.join(animaDir, "animaMaskImage")
+animaThrImage = os.path.join(animaDir, "animaThrImage")
+animaN4BiasCorrection = os.path.join(animaDir, "animaN4BiasCorrection")
+animaMorphologicalOperations = os.path.join(animaDir, "animaMorphologicalOperations")
+
+# Argument parsing
+parser = argparse.ArgumentParser(
+    description="Computes tissue segmentation from a prior atlas, multiple modalities inputs, and "
+                "EM tissue classification")
+
+parser.add_argument('-i', '--input', type=str, required=True, action='append',
+                    help='Image input: can give multiple modalities')
+parser.add_argument('-a', '--atlas', type=str, help='Atlas folder (default: use the one in anima scripts data '
+                                                    '- em_prior_atlas folder)')
+parser.add_argument('-m', '--mask', type=str, required=True, help='Brain mask of the first input')
+
+parser.add_argument('-o', '--output', type=str, help='Output path of the tissues segmentation '
+                                                     '(default is inputName_classification.nrrd)')
+parser.add_argument('-O', '--classes-output', type=str, help='Output path of the tissues probabilities')
+parser.add_argument('-z', '--zsc', type=str, help='Output path of the zscores of classification')
+parser.add_argument('-K', '--keep-intermediate-folder', action='store_true',
+                    help='Keep intermediate folder after script end')
+parser.add_argument('-P', '--prune-outliers', action='store_true',
+                    help='Remove outliers from the segmentation (according to z-score)')
+parser.add_argument('-Z', '--zsc-thr', type=int, default=4,
+                    help='Outliers threshold if -P is activated (default: 4)')
+
+args = parser.parse_args()
+
+numImages = len(sys.argv) - 1
+
+atlasDir = os.path.join(animaExtraDataDir,"em_prior_atlas")
+if args.atlas:
+    atlasDir = args.atlas
+
+atlasImage = os.path.join(atlasDir,"T1.nrrd")
+tissuesImage = os.path.join(atlasDir,"Tissue_Probabilities.nrrd")
+
+brainImages = args.input
+
+# Get floating image prefix
+brainImagePrefix = os.path.splitext(brainImages[0])[0]
+if os.path.splitext(brainImages[0])[1] == '.gz':
+    brainImagePrefix = os.path.splitext(brainImagePrefix)[0]
+
+tissuesOutputName = args.output if args.output else brainImagePrefix + "_classification.nrrd"
+intermediateFolder = os.path.join(os.path.dirname(brainImages[0]), 'em_tissues_' + str(uuid.uuid1()))
+
+if not os.path.isdir(intermediateFolder):
+    os.mkdir(intermediateFolder)
+
+brainImagePrefix = os.path.join(intermediateFolder, os.path.basename(brainImagePrefix))
+
+# Decide on whether to use large image setting or small image setting
+command = [animaConvertImage, "-i", brainImages[0], "-I"]
+convert_output = check_output(command, universal_newlines=True)
+size_info = convert_output.split('\n')[1].split('[')[1].split(']')[0]
+large_image = False
+for i in range(0, 3):
+    size_tmp = int(size_info.split(', ')[i])
+    if size_tmp >= 350:
+        large_image = True
+        break
+
+pyramidOptions = ["-p", "4", "-l", "1"]
+if large_image:
+    pyramidOptions = ["-p", "5", "-l", "2"]
+
+# First preprocess input images
+myfileImages = open(os.path.join(intermediateFolder,"listData.txt"),"w")
+for i in range(0,len(brainImages)):
+    currentData = brainImages[i]
+    if i > 0:
+        command = [animaPyramidalBMRegistration, "-m", currentData,
+                   "-r", brainImages[0],
+                   "-o", brainImagePrefix + "_rig_onRef_" + str(i) + ".nrrd", "--sp", "3"] + pyramidOptions
+        call(command)
+
+        currentData = brainImagePrefix + '_rig_onRef_' + str(i) + '.nrrd'
+
+    command = [animaMaskImage, "-i", currentData, "-m", args.mask,
+               "-o", brainImagePrefix + '_masked_' + str(i) + '.nrrd']
+    call(command)
+
+    command = [animaN4BiasCorrection, "-i", brainImagePrefix + '_masked_' + str(i) + '.nrrd',
+               "-o", brainImagePrefix + '_biasCorrected_' + str(i) + '.nrrd']
+    call(command)
+
+    myfileImages.write(brainImagePrefix + '_biasCorrected_' + str(i) + '.nrrd\n')
+
+myfileImages.close()
+
+# Now register prior atlas onto first reference image
+command = [animaPyramidalBMRegistration, "-m", atlasImage, "-r", brainImages[0], "-o", brainImagePrefix + "_rig.nrrd",
+           "-O", brainImagePrefix + "_rig_tr.txt", "--sp", "3"] + pyramidOptions
+call(command)
+
+command = [animaPyramidalBMRegistration, "-m", atlasImage, "-r", brainImages[0], "-o", brainImagePrefix + "_aff.nrrd",
+           "-O", brainImagePrefix + "_aff_tr.txt", "-i", brainImagePrefix + "_rig_tr.txt", "--sp", "3", "--ot",
+           "2"] + pyramidOptions
+call(command)
+
+command = [animaDenseSVFBMRegistration, "-r", brainImages[0], "-m", brainImagePrefix + "_aff.nrrd",
+           "-o", brainImagePrefix + "_nl.nrrd", "-O", brainImagePrefix + "_nl_tr.nrrd", "--tub", "2"] + pyramidOptions
+call(command)
+
+command = [animaTransformSerieXmlGenerator, "-i", brainImagePrefix + "_aff_tr.txt", "-i",
+           brainImagePrefix + "_nl_tr.nrrd", "-o", brainImagePrefix + "_nl_tr.xml"]
+call(command)
+
+command = [animaApplyTransformSerie, "-i", tissuesImage, "-t", brainImagePrefix + "_nl_tr.xml", "-g", brainImages[0],
+           "-o", brainImagePrefix + "_PriorTissues.nrrd"]
+call(command)
+
+# Finally, run classification step
+command = [animaTissuesEMClassification, "-i", os.path.join(intermediateFolder,"listData.txt"),
+           "-m", args.mask, "-t", brainImagePrefix + "_PriorTissues.nrrd", "-o", tissuesOutputName]
+
+if args.classes_output:
+    command = command + ["-O", args.classes_output]
+
+zscOut = args.zsc
+
+if not args.zsc and args.prune_outliers is True:
+    zscOut = brainImagePrefix + "_zsc.nrrd"
+
+if zscOut:
+    command = command + ["-z", zscOut]
+
+call(command)
+
+if args.prune_outliers is True:
+    command = [animaThrImage, "-i", zscOut, "-t", str(args.zsc_thr), "-o", brainImagePrefix + "_WrongMask.nrrd", "-I"]
+    call(command)
+
+    command = [animaMaskImage, "-i", tissuesOutputName, "-m", brainImagePrefix + "_WrongMask.nrrd",
+               "-o", tissuesOutputName]
+    call(command)
+
+if not args.keep_intermediate_folder:
+    rmtree(intermediateFolder)