-
Notifications
You must be signed in to change notification settings - Fork 7
/
IGR3D_MABMIS_Testing.cxx
1687 lines (1470 loc) · 70.4 KB
/
IGR3D_MABMIS_Testing.cxx
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*=========================================================================
Copyright (c) IDEA LAB, UNC-Chapel Hill, 2013.
MABMIS (Multi-Atlas-Based Multi-Image Segmentation)
This software is distributed WITHOUT ANY WARRANTY; without even
the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR
PURPOSE. See the above copyright notices for more information.
=========================================================================*/
// for math
#include <vcl_iostream.h>
#include <vnl/vnl_random.h>
#include <vnl/vnl_matrix.h>
#include <vnl/vnl_vector.h>
#include <vnl/vnl_matlab_print.h>
#include <vnl/algo/vnl_svd.h>
#include <vnl/algo/vnl_svd_economy.h>
#include <string>
#include <vector>
// basic itk
#include "itkVector.h"
// registration
#include "itkImageRegistrationMethod.h"
#include "itkSymmetricForcesDemonsRegistrationFilter.h"
// interpolator
#include "itkLinearInterpolateImageFunction.h"
#include "itkNearestNeighborInterpolateImageFunction.h"
// reader / writer
#include "itkImageFileReader.h"
#include "itkImageFileWriter.h"
// filter
#include "itkResampleImageFilter.h"
#include "itkCastImageFilter.h"
#include "itkWarpImageFilter.h"
#include "itkHistogramMatchingImageFilter.h"
#include "itkAddImageFilter.h"
#include "itkDivideImageFilter.h"
#include "itkMultiplyImageFilter.h"
#include "itkWarpVectorImageFilter.h"
#include "itkInverseDeformationFieldImageFilter.h"
// for affine transformation
#include "itkTransform.h"
#include "itkAffineTransform.h"
#include "itkImageRegistrationMethod.h"
// for Diffeomorphic Demons
#include <itkCommand.h>
#include <itkDiffeomorphicDemonsRegistrationFilter.h>
#include <itkMultiResolutionPDEDeformableRegistration.h>
// including itksys::SystemTools::MakeDirectory(char*)
#include <itksys/SystemTools.hxx>
#include <metaCommand.h>
// To include all related header files
#include "IGR3D_MABMIS_TestingCLP.h"
#include "itkMABMISImageOperationFilter.h"
#include "itkMABMISDeformationFieldFilter.h"
#include "itkMABMISSimulateData.h"
#include "itkMABMISImageRegistrationFilter.h"
#include "itkMABMISTreeOperation.h"
#include "itkMABMISBasicOperationFilter.h"
#include "itkMABMISAtlasXMLFile.h"
#include <algorithm>
static std::string ReplacePathSepForOS( const std::string & input )
{
std::string output = input;
#ifdef _WIN32
std::replace(output.begin(), output.end(), '/', FILESEP);
#else
std::replace(output.begin(), output.end(), '\\', FILESEP);
#endif
return output;
}
static std::string
GetExtension(const std::string & filename)
{
std::string fileExt( itksys::SystemTools::GetFilenameLastExtension(filename) );
//If the last extension is .gz, then need to pull off 2 extensions.
//.gz is the only valid compression extension.
if ( fileExt == std::string(".gz") )
{
fileExt = itksys::SystemTools::GetFilenameLastExtension(
itksys::SystemTools::GetFilenameWithoutLastExtension(filename) );
fileExt += ".gz";
}
return ( fileExt );
}
static std::string
GetRootName(const std::string & filename)
{
const std::string fileExt = GetExtension(filename);
// Create a base filename
// i.e Image.hdr --> Image
if ( fileExt.length() > 0 //Ensure that an extension was
// found
&& filename.length() > fileExt.length() //Ensure that the filename does
// not contain only the extension
)
{
const std::string::size_type it = filename.find_last_of(fileExt);
const std::string baseName( filename, 0, it - ( fileExt.length() - 1 ) );
return ( baseName );
}
//Default to return same as input when the extension is nothing (Analyze)
return ( filename );
}
typedef double CoordinateRepType;
const unsigned int SpaceDimension = ImageDimension;
// basic data type
typedef unsigned char CharPixelType; // for image IO usage
typedef float FloatPixelType; // for
typedef int IntPixelType;
typedef short ShortPixelType;
typedef float InternalPixelType; // for internal processing usage
typedef itk::Vector<InternalPixelType, ImageDimension> VectorPixelType;
// basic image type
typedef itk::Image<CharPixelType, ImageDimension> CharImageType;
typedef itk::Image<IntPixelType, ImageDimension> IntImageType;
typedef itk::Image<ShortPixelType, ImageDimension> ShortImageType;
typedef itk::Image<FloatPixelType, ImageDimension> FloatImageType;
typedef itk::Image<InternalPixelType, ImageDimension> InternalImageType;
typedef itk::Image<VectorPixelType, ImageDimension> DeformationFieldType;
// basic iterator type
typedef itk::ImageRegionIterator<DeformationFieldType> DeformationFieldIteratorType;
typedef itk::ImageRegionIterator<InternalImageType> InternalImageIteratorType;
typedef itk::ImageRegionIterator<CharImageType> CharImageIteratorType;
// basic image reader/writer related type
typedef itk::ImageFileReader<CharImageType> CharImageReaderType;
typedef itk::ImageFileReader<InternalImageType> InternalImageReaderType;
typedef itk::ImageFileWriter<InternalImageType> InternalImageWriterType;
typedef itk::WarpImageFilter<InternalImageType, InternalImageType, DeformationFieldType> InternalWarpFilterType;
typedef itk::ImageFileWriter<CharImageType> CharImageWriterType;
typedef itk::ImageFileWriter<IntImageType> IntImageWriterType;
typedef itk::ImageFileWriter<FloatImageType> FloatImageWriterType;
typedef itk::ImageFileWriter<ShortImageType> ShortImageWriterType;
typedef itk::ImageFileReader<DeformationFieldType> DeformationFieldReaderType;
typedef itk::ImageFileWriter<DeformationFieldType> DeformationFieldWriterType;
//////////////////////////////////////////////////////////////////////////////
// image filter type
typedef itk::ResampleImageFilter<InternalImageType, InternalImageType> ResampleFilterType;
typedef itk::HistogramMatchingImageFilter<InternalImageType, InternalImageType> InternalHistMatchFilterType;
////////////////////////////////////////////////////////////////////////////
// operation on deformation fields
typedef itk::WarpVectorImageFilter<DeformationFieldType, DeformationFieldType,
DeformationFieldType> WarpVectorFilterType;
typedef itk::InverseDeformationFieldImageFilter<DeformationFieldType,
DeformationFieldType>
InverseDeformationFieldImageFilterType;
typedef itk::AddImageFilter<DeformationFieldType, DeformationFieldType, DeformationFieldType> AddImageFilterType;
// global bool variables to adjust the procedure
bool isEvaluate = false; // if false, we do not know the ground-truth of labels
bool isDebug = false; // false;//true; // if true, print out more information
int localPatchSize = 1; // (2r+1)*(2r+1)*(2r+1) is the volume of local patch
// demons registration parameters
// int iterInResolutions[4][3]={{5,3,2},{10,5,5},{15,10,5},{20,15,10}};
// int itereach = 2; //
// int itereach0 = 0;int itereach1 = 1;int itereach2 = 2;int itereach3 = 3;
// double sigmaDef = 1.5;
// double sigmaDef10 = 1.0;double sigmaDef15 = 1.5;double sigmaDef20 = 2.0;
// double sigmaDef25 = 2.5;double sigmaDef30 = 3.0;double sigmaDef35 = 3.5;
bool doHistMatch = true;
bool doHistMatchTrue = true; bool doHistMatchFalse = false;
typedef itk::Statistics::MABMISSimulateData<InternalImageType, InternalImageType> DataSimulatorType;
DataSimulatorType::Pointer datasimulator;
typedef itk::Statistics::MABMISImageOperationFilter<CharImageType, CharImageType> ImageOperationFilterType;
ImageOperationFilterType::Pointer imgoperator;
typedef itk::Statistics::MABMISDeformationFieldFilter<InternalImageType,
InternalImageType> DeformationFieldOperationFilterType;
DeformationFieldOperationFilterType::Pointer dfoperator;
typedef itk::Statistics::MABMISImageRegistrationFilter<CharImageType, CharImageType> ImageRegistrationFilterType;
ImageRegistrationFilterType::Pointer regoperator;
typedef itk::Statistics::MABMISTreeOperation<InternalImageType, InternalImageType> TreeOperationType;
TreeOperationType::Pointer treeoperator;
typedef itk::Statistics::MABMISBasicOperationFilter<CharImageType, CharImageType> BasicOperationFilterType;
BasicOperationFilterType::Pointer basicoperator;
//
typedef itk::Vector<ShortPixelType, ImageDimension> ShortVectorPixelType;
typedef itk::Image<ShortVectorPixelType, ImageDimension> ShortDeformationFieldType;
typedef itk::ImageFileWriter<ShortDeformationFieldType> ShortDeformationFieldWriterType;
void ReadImgInfo(std::vector<std::string> imgfilenames);
void HistogramMatching(InternalImageType::Pointer inputImage, InternalImageType::Pointer referenceImage,
InternalImageType::Pointer & outputImage);
void TreeBasedRegistrationFastOniTree(vnl_vector<int> itree, // the incremental tree
int root, // the tree root
int itree_size, // the tree size
itk::MABMISAtlas* atlasTree, // the atlas tree
itk::MABMISImageData*imageData, // the images to be segmented
std::vector<int> iterations, // registration parameters
double sigma);
void LabelFusion(std::string curSampleImgName, std::string outSampleSegName, std::vector<std::string> allWarpedAtlasImgNames,
std::vector<std::string> allDeformationFieldNames, std::vector<std::string> allAtlasSegNames, int numOfAtlases);
void GaussianWeightedLabelFusion(InternalImageType::Pointer curSampleImgPtr, InternalImageType::Pointer outSampleSegPtr,
InternalImageType::Pointer* warpedImgPtrs, InternalImageType::Pointer* warpedSegPtrs,
int numOfAtlases);
void RegistrationOntoTreeRoot(vnl_vector<int> itree, // the incremental tree
int root, // the tree root
int itree_size, // the tree size
itk::MABMISAtlas* atlasTree, // the atlas tree
itk::MABMISImageData*imageData, // the images to be segmented
std::vector<int> iterations, // registration parameters
double sigma);
// void PairwiseRegistrationOnTreeViaRoot(int root, int filenumber, int atlas_size, std::vector<std::string> sub_ids);
void PairwiseRegistrationOnTreeViaRoot(int root, itk::MABMISImageData* imageData, itk::MABMISAtlas* atlasTree,
std::vector<int> iterations, double sigma);
// void MultiAtlasBasedSegmentation(int filenumber, int atlas_size, std::vector<std::string> sub_ids);
int MultiAtlasBasedSegmentation(int root, itk::MABMISImageData* imageData, itk::MABMISAtlas* atlasTree);
int Testing(itk::MABMISImageData* imageData, itk::MABMISAtlas* atlasTree,
std::vector<int> iterations, double sigma)
{
// Validate the tree size is correct
if( atlasTree->m_TreeSize != atlasTree->m_Tree.size() ||
atlasTree->m_TreeSize != atlasTree->m_NumberAllAtlases )
{
std::cerr << "ERROR: The atlas tree size is incorrect!!" << std::endl;
return -1;
}
/////////////////////////////////////
// the atlas tree
int tree_size = atlasTree->m_NumberAllAtlases;
// incremental tree: the atlas tree PLUS images to be segmented
int itree_size = atlasTree->m_NumberAllAtlases + imageData->m_NumberImageData;
vnl_vector<int> itree(itree_size);
for( int i = 0; i < itree_size; ++i )
{
itree.put(i, 0);
}
// copy current tree into an incremental tree
for( int i = 0; i < tree_size; ++i )
{
itree.put(i, atlasTree->m_Tree[i]); // combinative tree: all atlases including
// simulated data
}
treeoperator->FindRoot(itree, atlasTree->m_TreeSize, atlasTree->m_TreeRoot);
treeoperator->GetTreeHeight(itree, atlasTree->m_TreeSize, atlasTree->m_TreeHeight);
int root = atlasTree->m_TreeRoot; // root node
// Validate the tree is correct
for( int n = 0; n < atlasTree->m_TreeSize; ++n )
{
bool valid = true;
if( atlasTree->m_Tree[n] >= atlasTree->m_TreeSize )
{
valid = false;
}
if( atlasTree->m_Tree[n] < 0 && n != atlasTree->m_TreeRoot )
{
valid = false;
}
if( !valid )
{
std::cerr << "ERROR: The atlas tree size is incorrect!!" << std::endl;
return -1;
}
}
if( isDebug )
{
treeoperator->SaveTreeWithInfo(itree, itree_size, imageData->m_DataDirectory + "itree.txt");
}
// build the incremental tree
std::cout << "---------------------------------------" << std::endl;
std::cout << "Build incremental tree ... " << std::endl;
int totalNumAtlases = atlasTree->m_NumberAllAtlases;
int totalNumFiles = totalNumAtlases + imageData->m_NumberImageData;
std::vector<std::string> allfilenames(totalNumFiles);
for( int i = 0; i < totalNumFiles; ++i )
{
if( i < totalNumAtlases )
{
allfilenames[i] = ReplacePathSepForOS(atlasTree->m_AtlasDirectory + atlasTree->m_AtlasFilenames[i]);
}
else
{
allfilenames[i] = ReplacePathSepForOS(imageData->m_DataDirectory + imageData->m_ImageFileNames[i - totalNumAtlases]);
}
}
std::cout << "---------------------------------------" << std::endl;
std::cout << "1. Calculating pairwise distance cross training and test images ..." << std::endl;
vnl_matrix<double> cross_dist(imageData->m_NumberImageData, totalNumFiles);
for( int i = 0; i < imageData->m_NumberImageData; ++i )
{
for( int j = 0; j < totalNumFiles; ++j )
{
cross_dist.put(i, j, imgoperator->calculateDistanceMSD(allfilenames[totalNumAtlases + i], allfilenames[j]) );
}
}
if( isDebug )
{
basicoperator->SaveMatrix2File(cross_dist, imageData->m_NumberImageData, totalNumFiles,
imageData->m_DataDirectory + "sample2atlas_dist.txt");
}
// indices to for atlas files and image files to be segmented.
int* atlas_cur = new int[totalNumFiles];
for( int i = 0; i < totalNumFiles; ++i )
{
atlas_cur[i] = 0;
}
for( int i = 0; i < totalNumAtlases; ++i )
{
atlas_cur[i] = i;
}
int* images_cur = new int[imageData->m_NumberImageData];
for( int i = 0; i < imageData->m_NumberImageData; ++i )
{
images_cur[i] = i + totalNumAtlases;
}
// build the incremental tree
treeoperator->SetAllAtlasNumber(totalNumAtlases);
std::cout << "---------------------------------------" << std::endl;
std::cout << "2. Build the tree... " << std::endl;
itree = treeoperator->BuildIncrementalTree(imageData->m_NumberImageData, // #images to be segmented
totalNumAtlases, // #atlases, including original ones and
// simulated ones
totalNumFiles, // total #image files, =
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
// imageData->m_NumberImageData+totalNumAtlases
images_cur, // file indices for images to be segmented
atlas_cur, // file indices for atlas images
cross_dist, // pairwise distance between images
itree); // the output incremental tree
// delete
delete[] atlas_cur;
delete[] images_cur;
// check if new tree is actually a tree
const bool iTreeValid = treeoperator->ValidateTree(itree, itree_size);
if( !iTreeValid )
{
std::cout << "The new built tree is NOT valid!" << std::endl;
return 105; // EXIT_FAILURE;
}
// save new tree
if( isDebug )
{
treeoperator->SaveTreeWithInfo(itree, itree_size, imageData->m_DataDirectory + "itree.txt");
// std::cout << "Tree is built up!" << std::endl;
}
std::cout << "Done! " << std::endl;
std::cout << "---------------------------------------" << std::endl;
std::cout << "3. Register all images to the root... " << std::endl;
// register all images to the root, if not done yet.
RegistrationOntoTreeRoot(itree, // the incremental tree
root, // the tree root
itree_size, // the tree size
atlasTree,
imageData,
iterations, sigma // registration parameters
);
std::cout << "Done. " << std::endl;
std::cout << "---------------------------------------" << std::endl;
std::cout << "4. Pairwise registration via the tree root... " << std::endl;
PairwiseRegistrationOnTreeViaRoot(root, imageData, atlasTree, iterations, sigma);
std::cout << "Done. " << std::endl;
////////////////////////////////
// do all multi-atlas based segmentation with weighted label fusion
std::cout << "---------------------------------------" << std::endl;
std::cout << "5. Segment images using label fusion... " << std::endl;
int numIter = MultiAtlasBasedSegmentation(root, imageData, atlasTree);
std::cout << "---------------------------------------" << std::endl;
std::cout << "Done. " << std::endl;
// do all multi-atlas based segmentation with weighted label fusion
/////////////////////////////////
//
// basicoperator->RemoveFile("*_*_cbq_000*");
// basicoperator->RemoveFile("*_*_deform_000*");
//
/////////////////////////////////
// remove intermediate results
// deformedImageFileName and deformation fields
if( !isDebug )
{
std::cout << "---------------------------------------" << std::endl;
std::cout << "Removing intermediate registration files..." << std::endl;
for( int n = 0; n < imageData->m_NumberImageData; ++n )
{
char n_str[10], m_str[10];
sprintf(n_str, "%03d", n);
for( int m = 0; m < atlasTree->m_NumberAllAtlases - atlasTree->m_NumberSimulatedAtlases; ++m )
{
sprintf(m_str, "%03d", m);
std::string imHdr = std::string(m_str) + "_to_" + "testImage" + n_str + "_cbq_000.nii.gz";
std::string dfMha = std::string(m_str) + "_to_" + "testImage" + n_str + "_deform_000.nii.gz";
imHdr = imageData->m_DataDirectory + imHdr;
dfMha = imageData->m_DataDirectory + dfMha;
basicoperator->RemoveFile(imHdr.c_str() );
basicoperator->RemoveFile(dfMha.c_str() );
if( m == root )
{
imHdr = std::string("testImage") + n_str + "_to_" + m_str + "_cbq_reg.nii.gz";
dfMha = std::string("testImage") + n_str + "_to_" + m_str + "_deform_000.nii.gz";
imHdr = imageData->m_DataDirectory + imHdr;
dfMha = imageData->m_DataDirectory + dfMha;
basicoperator->RemoveFile(imHdr.c_str() );
basicoperator->RemoveFile(dfMha.c_str() );
}
}
for( int m = 0; m < imageData->m_NumberImageData; ++m )
{
if( m == n )
{
continue;
}
sprintf(m_str, "%03d", m);
std::string imHdr = std::string("testImage") + n_str + "_to_testImage" + m_str + "_cbq_000.nii.gz";
std::string dfMha = std::string("testImage") + n_str + "_to_testImage" + m_str + "_deform_000.nii.gz";
imHdr = imageData->m_DataDirectory + imHdr;
dfMha = imageData->m_DataDirectory + dfMha;
basicoperator->RemoveFile(imHdr.c_str() );
basicoperator->RemoveFile(dfMha.c_str() );
}
}
}
// remove segmentation results in iterations.
for( int n = 0; n < imageData->m_NumberImageData; ++n )
{
const std::string imageFileName = imageData->m_DataDirectory + imageData->m_ImageFileNames[n];
const std::string baseFileName = GetRootName(imageFileName);
char i_str[10];
bool copied = false;
for( int i = numIter - 1; i >= 0; i-- )
{
sprintf(i_str, "%03d", i);
const std::string segHdr = baseFileName + "_seg_" + i_str + ".nii.gz";
if( itksys::SystemTools::FileExists(segHdr.c_str(), true) )
{
if( !copied )
{
std::string segHdr_save = baseFileName + "_seg.nii.gz";
const size_t dir_sep = segHdr_save.find_last_of(FILESEP);
if( dir_sep != std::string::npos )
{
segHdr_save = imageData->m_OutputDirectory + segHdr_save.substr(dir_sep + 1, std::string::npos);
}
else
{
segHdr_save = imageData->m_OutputDirectory + segHdr_save;
}
itksys::SystemTools::CopyFileAlways(segHdr.c_str(), segHdr_save.c_str() );
copied = true;
}
basicoperator->RemoveFile(segHdr.c_str() );
}
}
}
std::cout << "Done. " << std::endl;
return EXIT_SUCCESS;
}
template <class T>
int DoIt( itk::MABMISImageData* imageData, itk::MABMISAtlas* atlasTree,
std::vector<int> iterations, double sigma)
{
return Testing(imageData, atlasTree, iterations, sigma);
}
int main( int argc, char *argv[] )
{
PARSE_ARGS;
// step 1: read the test image list
itk::MABMISImageDataXMLFileReader::Pointer imageListXMLReader = itk::MABMISImageDataXMLFileReader::New();
ImageListXML = ReplacePathSepForOS(ImageListXML);
AtlaseTreeXML = ReplacePathSepForOS(AtlaseTreeXML);
OutputFolder = ReplacePathSepForOS(OutputFolder);
imageListXMLReader->SetFilename(ImageListXML);
try
{
imageListXMLReader->GenerateOutputInformation();
}
catch( itk::ExceptionObject & excp )
{
std::cerr << "Exception thrown while reading file" << std::endl;
std::cerr << excp << std::endl;
return EXIT_FAILURE;
}
itk::MABMISImageData * inputImageData = imageListXMLReader->GetOutputObject();
// if the data path is empty, use the path of the xml file instead
if( inputImageData->m_DataDirectory.size() <= 1 )
{
const size_t dir_sep = ImageListXML.find_last_of(FILESEP);
if( dir_sep != std::string::npos )
{
inputImageData->m_DataDirectory = ImageListXML.substr(0, dir_sep);
}
else
{
inputImageData->m_DataDirectory.resize(0);
}
}
if( !inputImageData->m_DataDirectory.empty() )
{
if( !(inputImageData->m_DataDirectory[inputImageData->m_DataDirectory.size() - 1] == FILESEP) )
{
inputImageData->m_DataDirectory = inputImageData->m_DataDirectory + FILESEP;
}
}
// output directory
if( OutputFolder.empty() )
{
inputImageData->m_OutputDirectory = inputImageData->m_DataDirectory;
}
else
{
if( !(OutputFolder[OutputFolder.size() - 1] == FILESEP) )
{
OutputFolder = OutputFolder + FILESEP;
}
inputImageData->m_OutputDirectory = OutputFolder;
}
// load the tree-structured atlases that is generated in the training step
itk::MABMISAtlasXMLFileReader::Pointer treeAtlasXMLReader = itk::MABMISAtlasXMLFileReader::New();
treeAtlasXMLReader->SetFilename(AtlaseTreeXML);
try
{
treeAtlasXMLReader->GenerateOutputInformation();
}
catch( itk::ExceptionObject & excp )
{
std::cerr << "Exception thrown while reading file" << std::endl;
std::cerr << excp << std::endl;
return EXIT_FAILURE;
}
itk::MABMISAtlas * atlasTree = treeAtlasXMLReader->GetOutputObject();
// set the atlas path as the same path as the xml file.
const size_t dir_sep = AtlaseTreeXML.find_last_of(FILESEP);
if( dir_sep != std::string::npos )
{
atlasTree->m_AtlasDirectory = ReplacePathSepForOS(AtlaseTreeXML.substr(0, dir_sep + 1) + atlasTree->m_AtlasDirectory);
}
if( atlasTree->m_AtlasDirectory.size() == 0 )
{
atlasTree->m_AtlasDirectory = ReplacePathSepForOS(".");
}
if( !(atlasTree->m_AtlasDirectory[atlasTree->m_AtlasDirectory.size() - 1] == FILESEP) )
{
atlasTree->m_AtlasDirectory = ReplacePathSepForOS(atlasTree->m_AtlasDirectory + FILESEP );
}
// Will look into getting rid of these global variables later ---Xiaofeng
datasimulator = DataSimulatorType::New();
imgoperator = ImageOperationFilterType::New();
dfoperator = DeformationFieldOperationFilterType::New();
regoperator = ImageRegistrationFilterType::New();
treeoperator = TreeOperationType::New();
basicoperator = BasicOperationFilterType::New();
int retVal = DoIt<unsigned short>( inputImageData, atlasTree, iterations, SmoothingKernelSize);
delete inputImageData;
delete atlasTree;
return retVal;
}
void LabelFusion(std::string curSampleImgName, std::string outSampleSegName, std::vector<std::string> allWarpedAtlasImgNames,
std::vector<std::string> allDeformationFieldNames, std::vector<std::string> allAtlasSegNames, int numOfAtlases)
{
// create pointers for each images including: cur_image, our_seg,
// warpedImg(after histogram matching), warpedSeg (after warping)
// cur sample image pointer
InternalImageType::Pointer curSampleImgPtr = 0;
imgoperator->ReadImage(curSampleImgName, curSampleImgPtr);
// output sample label image pointer
InternalImageType::Pointer outSampleSegPtr = 0;
imgoperator->ReadImage(curSampleImgName, outSampleSegPtr);
// atlas related images pointers
InternalImageType::Pointer* warpedImgPtrs = new InternalImageType::Pointer[numOfAtlases];
InternalImageType::Pointer* warpedSegPtrs = new InternalImageType::Pointer[numOfAtlases];
InternalImageType::IndexType index;
index[0] = 19; index[1] = 19; index[2] = 8;
for( int i = 0; i < numOfAtlases; ++i )
{
// histogram match each warped atlas to the current sample
warpedImgPtrs[i] = 0;
InternalImageType::Pointer curWarpedAtlasPtr = 0;
imgoperator->ReadImage(allWarpedAtlasImgNames[i], curWarpedAtlasPtr);
HistogramMatching(curWarpedAtlasPtr, curSampleImgPtr, warpedImgPtrs[i]);
// load each deformation field
DeformationFieldType::Pointer deformFieldPtr = 0;
dfoperator->ReadDeformationField(allDeformationFieldNames[i], deformFieldPtr);
// warp each atlas label image to the current sample space
InternalImageType::Pointer curAtlasSegPtr = 0;
imgoperator->ReadImage(allAtlasSegNames[i], curAtlasSegPtr);
dfoperator->ApplyDeformationField(curAtlasSegPtr, deformFieldPtr, warpedSegPtrs[i], false);
}
// weighted label fusion
GaussianWeightedLabelFusion(curSampleImgPtr, outSampleSegPtr, warpedImgPtrs, warpedSegPtrs, numOfAtlases);
// output segmentation image
imgoperator->WriteImage(outSampleSegName, outSampleSegPtr);
delete[] warpedImgPtrs;
delete[] warpedSegPtrs;
return;
}
void GaussianWeightedLabelFusion(InternalImageType::Pointer curSampleImgPtr, InternalImageType::Pointer outSampleSegPtr,
InternalImageType::Pointer* warpedImgPtrs, InternalImageType::Pointer* warpedSegPtrs,
int numOfAtlases)
{
// introduce the iterators of each image
// InternalImageIteratorType sampleImgIt(curSampleImgPtr,
// curSampleImgPtr->GetLargestPossibleRegion());//GetRequestedRegion());
InternalImageIteratorType sampleSegIt(outSampleSegPtr, outSampleSegPtr->GetLargestPossibleRegion() ); //
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
// GetRequestedRegion());
InternalImageType::SizeType sampleImSize = outSampleSegPtr->GetBufferedRegion().GetSize();
// InternalImageIteratorType* warpedImgIts = new InternalImageIteratorType[numOfAtlases];
InternalImageIteratorType* warpedSegIts = new InternalImageIteratorType[numOfAtlases];
typedef itk::ConstNeighborhoodIterator<InternalImageType> InternalImageNeighborhoodIteratorType;
InternalImageNeighborhoodIteratorType* warpedImgNeighborhoodIts =
new InternalImageNeighborhoodIteratorType[numOfAtlases];
InternalImageType::SizeType radius;
radius[0] = localPatchSize; radius[1] = localPatchSize; radius[2] = localPatchSize;
InternalImageNeighborhoodIteratorType sampleImgNeighborhoodIt(radius, curSampleImgPtr,
curSampleImgPtr->GetRequestedRegion() );
InternalImageNeighborhoodIteratorType sampleImgNeighborIt(radius, curSampleImgPtr,
curSampleImgPtr->GetLargestPossibleRegion() );
for( int i = 0; i < numOfAtlases; ++i )
{
// InternalImageIteratorType it1( warpedImgPtrs[i],
// warpedImgPtrs[i]->GetLargestPossibleRegion());//GetRequestedRegion() );
// warpedImgIts[i] = it1;
InternalImageIteratorType it2( warpedSegPtrs[i], warpedSegPtrs[i]->GetLargestPossibleRegion() ); //
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
//
// GetRequestedRegion()
// );
warpedSegIts[i] = it2;
InternalImageNeighborhoodIteratorType neighborIt(radius, warpedImgPtrs[i],
warpedImgPtrs[i]->GetLargestPossibleRegion() );
warpedImgNeighborhoodIts[i] = neighborIt;
}
int maxLabel = 0; // the maximum labels
// find out the maximum label from the label images
for( int i = 0; i < numOfAtlases; ++i )
{
warpedSegIts[i].GoToBegin();
}
while( !warpedSegIts[0].IsAtEnd() )
{
for( int i = 0; i < numOfAtlases; ++i )
{
if( warpedSegIts[i].Get() > maxLabel )
{
maxLabel = warpedSegIts[i].Get();
}
++warpedSegIts[i];
}
}
InternalImageIteratorType::IndexType idx;
InternalImageIteratorType::IndexType idx1;
int temp1, temp2;
double kernel_sigma = 1; // std of Gaussian approximation, set to the median of all distances
int* index = new int[numOfAtlases];
double* sort1 = new double[numOfAtlases];
int* label_index = new int[maxLabel + 1];
int* label_pool = new int[numOfAtlases];
double* mse = new double[numOfAtlases];
double* weight_sum = new double[maxLabel + 1];
// for each voxel do label fusion,(vx, vy, vz) is the current location
double kernel_sigma_square = kernel_sigma * kernel_sigma;
for( int i = 0; i < numOfAtlases; ++i )
{
warpedImgNeighborhoodIts[i].GoToBegin();
warpedSegIts[i].GoToBegin();
}
for( sampleImgNeighborIt.GoToBegin(), sampleSegIt.GoToBegin();
!sampleImgNeighborIt.IsAtEnd();
++sampleImgNeighborIt, ++sampleSegIt )
{
// idx = sampleImgNeighborIt.GetIndex();
// if (idx[0] == 30 && idx[1]==20 && idx[2] ==8)
// idx[0] = idx[0];
double msec = 0.0;
for( int i = 0; i < numOfAtlases; ++i )
{
label_pool[i] = warpedSegIts[i].Get();
for( int n = 0; n < sampleImgNeighborIt.Size(); ++n )
{
temp1 = sampleImgNeighborIt.GetPixel(n);
temp2 = warpedImgNeighborhoodIts[i].GetPixel(n);
// add together differences, squared sum
msec += (temp1 - temp2) * (temp1 - temp2);
}
mse[i] = msec;
}
// sort the mean square difference
for( int i = 0; i < numOfAtlases; ++i )
{
index[i] = i;
sort1[i] = mse[i];
}
basicoperator->bubbleSort(sort1, index, numOfAtlases);
kernel_sigma_square = sort1[(int)( (numOfAtlases - 1) / 2)] + 0.0001;
// weight each label
for( int i = 0; i < maxLabel + 1; ++i )
{
weight_sum[i] = 0.0;
}
for( int i = 0; i < numOfAtlases; ++i )
{
// add-up all weights
if ((mse[i]) / (2.0 * kernel_sigma_square) < 50)
weight_sum[label_pool[i]] += exp(0.0 - (mse[i]) / (2.0 * kernel_sigma_square) );
}
// weighted label fusion
for( int i = 0; i < maxLabel + 1; ++i )
{
label_index[i] = i;
}
// determine the label with the maximum weight
// basicoperator->bubbleSort(weight_sum, label_index, maxLabel+1);
// int label_final;
// label_final = label_index[maxLabel];
int label_final = 0;
double weight_final = 0.0;
for( int i = 0; i < maxLabel + 1; ++i )
{
if( weight_sum[i] > weight_final )
{
label_final = i;
weight_final = weight_sum[i];
}
}
// assign label to the segmentations
sampleSegIt.Set(label_final);
for( int i = 0; i < numOfAtlases; ++i )
{
++warpedImgNeighborhoodIts[i];
++warpedSegIts[i];
}
}
/*
for (int vz = 0; vz < sampleImSize[2]; vz++)
{
//if (isDebug)
// std::cout << vz << ", ";
for (int vy = 0; vy < sampleImSize[1]; vy++)
{
for (int vx = 0; vx < sampleImSize[0]; vx++)
{
// process each voxel
idx.SetElement(0,vx);
idx.SetElement(1,vy);
idx.SetElement(2,vz);
// get labels from all atlases on this current location
for (int i = 0; i < numOfAtlases; ++i)
{
warpedSegIts[i].SetIndex(idx);
label_pool[i] = warpedSegIts[i].Get();
}
// get the local patch differences
for (int i = 0; i < numOfAtlases; ++i) mse[i] = 0.0;
for (int i = 0; i < numOfAtlases; ++i)
{
double msec = 0.0;
// compare with all other images at the local patch of current voxel
// (nx,ny,nz) is the incremental position on (vx,vy,vz)
for(int nz = 0-localPatchSize; nz <= localPatchSize; nz++)
{
for(int ny = 0-localPatchSize; ny <= localPatchSize; ny++)
{
for(int nx = 0-localPatchSize; nx <= localPatchSize; nx++)
{
if ((vx+nx>=0)&&(vx+nx<sampleImSize[0])&&(vy+ny>=0)&&(vy+ny<sampleImSize[1])&&(vz+nz>=0)&&(vz+nz<sampleImSize[2]))
{
// within the valid range, and then get voxel intensity
//
idx1.SetElement(0,vx+nx);
idx1.SetElement(1,vy+ny);
idx1.SetElement(2,vz+nz);
sampleImgIt.SetIndex(idx1);
warpedImgIts[i].SetIndex(idx1);
temp1 = sampleImgIt.Get();
temp2 = warpedImgIts[i].Get();
// add together differences, squared sum
msec += (temp1-temp2)*(temp1-temp2);
}
}
}
}
mse[i] = sqrt(msec);
}// end of for nz
// sort all difference
for (int i = 0; i < numOfAtlases; ++i)
{
index[i] = i;
sort1[i] = mse[i];
}
basicoperator->bubbleSort(sort1, index, numOfAtlases);
kernel_sigma = sort1[(int)((numOfAtlases-1)/2)]+0.0001;
// weight each label
for (int i = 0; i < maxLabel+1; ++i) weight_sum[i] = 0.0;
for (int i = 0; i < numOfAtlases; ++i)
{
// add-up all weights
weight_sum[label_pool[i]]+= exp(0-(mse[i]*mse[i])/(2*kernel_sigma*kernel_sigma));
}
// weighted label fusion
for (int i = 0; i < maxLabel+1; ++i) label_index[i] = i;
basicoperator->bubbleSort(weight_sum, label_index, maxLabel+1);
int label_final;
label_final = label_index[maxLabel];
// assign label to the segmentations
sampleSegIt.SetIndex(idx);
sampleSegIt.Set(label_final);
}
}
}// end of for vz
*/
delete[] label_pool;
delete[] mse;
delete[] weight_sum;
//
// delete[] warpedImgIts;
delete[] warpedSegIts;
delete[] warpedImgNeighborhoodIts;
delete[] sort1;
delete[] index;
delete[] label_index;
return;
}
void TreeBasedRegistrationFastOniTree(vnl_vector<int> itree, // the incremental tree
int root, // the tree root
int itree_size, // the tree size
itk::MABMISAtlas* atlasTree, // the atlas tree
itk::MABMISImageData*imageData, // the images to be segmented
std::vector<int> iterations,
double sigma
)
// (vnl_vector<int> itree_t, int itree_size_t, int atlas_size_t, int simulate_size_t, int sample_size_t, std::vector<std::string>
// originalIntensityImageFileNames, std::vector<std::string> deformedImgFileNames, std::vector<std::string> deformationFieldFileNames)
{
// int root = -1; //
// treeoperator->FindRoot(itree_t, itree_size_t, root);
// calculate the depth of each node
int* node_depth = new int[itree_size];