-
Notifications
You must be signed in to change notification settings - Fork 3
/
nifti1.h
1490 lines (1228 loc) · 67.6 KB
/
nifti1.h
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
/** \file nifti1.h
\brief Official definition of the nifti1 header. Written by Bob Cox, SSCC, NIMH.
HISTORY:
29 Nov 2007 [rickr]
- added DT_RGBA32 and NIFTI_TYPE_RGBA32
- added NIFTI_INTENT codes:
TIME_SERIES, NODE_INDEX, RGB_VECTOR, RGBA_VECTOR, SHAPE
*/
#ifndef _NIFTI_HEADER_
#define _NIFTI_HEADER_
/*****************************************************************************
** This file defines the "NIFTI-1" header format. **
** It is derived from 2 meetings at the NIH (31 Mar 2003 and **
** 02 Sep 2003) of the Data Format Working Group (DFWG), **
** chartered by the NIfTI (Neuroimaging Informatics Technology **
** Initiative) at the National Institutes of Health (NIH). **
**--------------------------------------------------------------**
** Neither the National Institutes of Health (NIH), the DFWG, **
** nor any of the members or employees of these institutions **
** imply any warranty of usefulness of this material for any **
** purpose, and do not assume any liability for damages, **
** incidental or otherwise, caused by any use of this document. **
** If these conditions are not acceptable, do not use this! **
**--------------------------------------------------------------**
** Author: Robert W Cox (NIMH, Bethesda) **
** Advisors: John Ashburner (FIL, London), **
** Stephen Smith (FMRIB, Oxford), **
** Mark Jenkinson (FMRIB, Oxford) **
******************************************************************************/
/*---------------------------------------------------------------------------*/
/* Note that the ANALYZE 7.5 file header (dbh.h) is
(c) Copyright 1986-1995
Biomedical Imaging Resource
Mayo Foundation
Incorporation of components of dbh.h are by permission of the
Mayo Foundation.
Changes from the ANALYZE 7.5 file header in this file are released to the
public domain, including the functional comments and any amusing asides.
-----------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/*! INTRODUCTION TO NIFTI-1:
------------------------
The twin (and somewhat conflicting) goals of this modified ANALYZE 7.5
format are:
(a) To add information to the header that will be useful for functional
neuroimaging data analysis and display. These additions include:
- More basic data types.
- Two affine transformations to specify voxel coordinates.
- "Intent" codes and parameters to describe the meaning of the data.
- Affine scaling of the stored data values to their "true" values.
- Optional storage of the header and image data in one file (.nii).
(b) To maintain compatibility with non-NIFTI-aware ANALYZE 7.5 compatible
software (i.e., such a program should be able to do something useful
with a NIFTI-1 dataset -- at least, with one stored in a traditional
.img/.hdr file pair).
Most of the unused fields in the ANALYZE 7.5 header have been taken,
and some of the lesser-used fields have been co-opted for other purposes.
Notably, most of the data_history substructure has been co-opted for
other purposes, since the ANALYZE 7.5 format describes this substructure
as "not required".
NIFTI-1 FLAG (MAGIC STRINGS):
----------------------------
To flag such a struct as being conformant to the NIFTI-1 spec, the last 4
bytes of the header must be either the C String "ni1" or "n+1";
in hexadecimal, the 4 bytes
6E 69 31 00 or 6E 2B 31 00
(in any future version of this format, the '1' will be upgraded to '2',
etc.). Normally, such a "magic number" or flag goes at the start of the
file, but trying to avoid clobbering widely-used ANALYZE 7.5 fields led to
putting this marker last. However, recall that "the last shall be first"
(Matthew 20:16).
If a NIFTI-aware program reads a header file that is NOT marked with a
NIFTI magic string, then it should treat the header as an ANALYZE 7.5
structure.
NIFTI-1 FILE STORAGE:
--------------------
"ni1" means that the image data is stored in the ".img" file corresponding
to the header file (starting at file offset 0).
"n+1" means that the image data is stored in the same file as the header
information. We recommend that the combined header+data filename suffix
be ".nii". When the dataset is stored in one file, the first byte of image
data is stored at byte location (int)vox_offset in this combined file.
The minimum allowed value of vox_offset is 352; for compatibility with
some software, vox_offset should be an integral multiple of 16.
GRACE UNDER FIRE:
----------------
Most NIFTI-aware programs will only be able to handle a subset of the full
range of datasets possible with this format. All NIFTI-aware programs
should take care to check if an input dataset conforms to the program's
needs and expectations (e.g., check datatype, intent_code, etc.). If the
input dataset can't be handled by the program, the program should fail
gracefully (e.g., print a useful warning; not crash).
SAMPLE CODES:
------------
The associated files nifti1_io.h and nifti1_io.c provide a sample
implementation in C of a set of functions to read, write, and manipulate
NIFTI-1 files. The file nifti1_test.c is a sample program that uses
the nifti1_io.c functions.
-----------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* HEADER STRUCT DECLARATION:
-------------------------
In the comments below for each field, only NIFTI-1 specific requirements
or changes from the ANALYZE 7.5 format are described. For convenience,
the 348 byte header is described as a single struct, rather than as the
ANALYZE 7.5 group of 3 substructs.
Further comments about the interpretation of various elements of this
header are after the data type definition itself. Fields that are
marked as ++UNUSED++ have no particular interpretation in this standard.
(Also see the UNUSED FIELDS comment section, far below.)
The presumption below is that the various C types have particular sizes:
sizeof(int) = sizeof(float) = 4 ; sizeof(short) = 2
-----------------------------------------------------------------------------*/
/*=================*/
#ifdef __cplusplus
extern "C" {
#endif
/*=================*/
/*! \struct nifti_1_header
\brief Data structure defining the fields in the nifti1 header.
This binary header should be found at the beginning of a valid
NIFTI-1 header file.
*/
/*************************/ /************************/
struct nifti_1_header { /* NIFTI-1 usage */ /* ANALYZE 7.5 field(s) */
/*************************/ /************************/
/*--- was header_key substruct ---*/
int sizeof_hdr; /*!< MUST be 348 */ /* int sizeof_hdr; */
char data_type[10]; /*!< ++UNUSED++ */ /* char data_type[10]; */
char db_name[18]; /*!< ++UNUSED++ */ /* char db_name[18]; */
int extents; /*!< ++UNUSED++ */ /* int extents; */
short session_error; /*!< ++UNUSED++ */ /* short session_error; */
char regular; /*!< ++UNUSED++ */ /* char regular; */
char dim_info; /*!< MRI slice ordering. */ /* char hkey_un0; */
/*--- was image_dimension substruct ---*/
short dim[8]; /*!< Data array dimensions.*/ /* short dim[8]; */
float intent_p1 ; /*!< 1st intent parameter. */ /* short unused8; */
/* short unused9; */
float intent_p2 ; /*!< 2nd intent parameter. */ /* short unused10; */
/* short unused11; */
float intent_p3 ; /*!< 3rd intent parameter. */ /* short unused12; */
/* short unused13; */
short intent_code ; /*!< NIFTI_INTENT_* code. */ /* short unused14; */
short datatype; /*!< Defines data type! */ /* short datatype; */
short bitpix; /*!< Number bits/voxel. */ /* short bitpix; */
short slice_start; /*!< First slice index. */ /* short dim_un0; */
float pixdim[8]; /*!< Grid spacings. */ /* float pixdim[8]; */
float vox_offset; /*!< Offset into .nii file */ /* float vox_offset; */
float scl_slope ; /*!< Data scaling: slope. */ /* float funused1; */
float scl_inter ; /*!< Data scaling: offset. */ /* float funused2; */
short slice_end; /*!< Last slice index. */ /* float funused3; */
char slice_code ; /*!< Slice timing order. */
char xyzt_units ; /*!< Units of pixdim[1..4] */
float cal_max; /*!< Max display intensity */ /* float cal_max; */
float cal_min; /*!< Min display intensity */ /* float cal_min; */
float slice_duration;/*!< Time for 1 slice. */ /* float compressed; */
float toffset; /*!< Time axis shift. */ /* float verified; */
int glmax; /*!< ++UNUSED++ */ /* int glmax; */
int glmin; /*!< ++UNUSED++ */ /* int glmin; */
/*--- was data_history substruct ---*/
char descrip[80]; /*!< any text you like. */ /* char descrip[80]; */
char aux_file[24]; /*!< auxiliary filename. */ /* char aux_file[24]; */
short qform_code ; /*!< NIFTI_XFORM_* code. */ /*-- all ANALYZE 7.5 ---*/
short sform_code ; /*!< NIFTI_XFORM_* code. */ /* fields below here */
/* are replaced */
float quatern_b ; /*!< Quaternion b param. */
float quatern_c ; /*!< Quaternion c param. */
float quatern_d ; /*!< Quaternion d param. */
float qoffset_x ; /*!< Quaternion x shift. */
float qoffset_y ; /*!< Quaternion y shift. */
float qoffset_z ; /*!< Quaternion z shift. */
float srow_x[4] ; /*!< 1st row affine transform. */
float srow_y[4] ; /*!< 2nd row affine transform. */
float srow_z[4] ; /*!< 3rd row affine transform. */
char intent_name[16];/*!< 'name' or meaning of data. */
char magic[4] ; /*!< MUST be "ni1\0" or "n+1\0". */
} ; /**** 348 bytes total ****/
typedef struct nifti_1_header nifti_1_header ;
/*---------------------------------------------------------------------------*/
/* HEADER EXTENSIONS:
-----------------
After the end of the 348 byte header (e.g., after the magic field),
the next 4 bytes are a char array field named "extension". By default,
all 4 bytes of this array should be set to zero. In a .nii file, these
4 bytes will always be present, since the earliest start point for
the image data is byte #352. In a separate .hdr file, these bytes may
or may not be present. If not present (i.e., if the length of the .hdr
file is 348 bytes), then a NIfTI-1 compliant program should use the
default value of extension={0,0,0,0}. The first byte (extension[0])
is the only value of this array that is specified at present. The other
3 bytes are reserved for future use.
If extension[0] is nonzero, it indicates that extended header information
is present in the bytes following the extension array. In a .nii file,
this extended header data is before the image data (and vox_offset
must be set correctly to allow for this). In a .hdr file, this extended
data follows extension and proceeds (potentially) to the end of the file.
The format of extended header data is weakly specified. Each extension
must be an integer multiple of 16 bytes long. The first 8 bytes of each
extension comprise 2 integers:
int esize , ecode ;
These values may need to be byte-swapped, as indicated by dim[0] for
the rest of the header.
* esize is the number of bytes that form the extended header data
+ esize must be a positive integral multiple of 16
+ this length includes the 8 bytes of esize and ecode themselves
* ecode is a non-negative integer that indicates the format of the
extended header data that follows
+ different ecode values are assigned to different developer groups
+ at present, the "registered" values for code are
= 0 = unknown private format (not recommended!)
= 2 = DICOM format (i.e., attribute tags and values)
= 4 = AFNI group (i.e., ASCII XML-ish elements)
In the interests of interoperability (a primary rationale for NIfTI),
groups developing software that uses this extension mechanism are
encouraged to document and publicize the format of their extensions.
To this end, the NIfTI DFWG will assign even numbered codes upon request
to groups submitting at least rudimentary documentation for the format
of their extension; at present, the contact is mailto:rwcox@nih.gov.
The assigned codes and documentation will be posted on the NIfTI
website. All odd values of ecode (and 0) will remain unassigned;
at least, until the even ones are used up, when we get to 2,147,483,646.
Note that the other contents of the extended header data section are
totally unspecified by the NIfTI-1 standard. In particular, if binary
data is stored in such a section, its byte order is not necessarily
the same as that given by examining dim[0]; it is incumbent on the
programs dealing with such data to determine the byte order of binary
extended header data.
Multiple extended header sections are allowed, each starting with an
esize,ecode value pair. The first esize value, as described above,
is at bytes #352-355 in the .hdr or .nii file (files start at byte #0).
If this value is positive, then the second (esize2) will be found
starting at byte #352+esize1 , the third (esize3) at byte #352+esize1+esize2,
et cetera. Of course, in a .nii file, the value of vox_offset must
be compatible with these extensions. If a malformed file indicates
that an extended header data section would run past vox_offset, then
the entire extended header section should be ignored. In a .hdr file,
if an extended header data section would run past the end-of-file,
that extended header data should also be ignored.
With the above scheme, a program can successively examine the esize
and ecode values, and skip over each extended header section if the
program doesn't know how to interpret the data within. Of course, any
program can simply ignore all extended header sections simply by jumping
straight to the image data using vox_offset.
-----------------------------------------------------------------------------*/
/*! \struct nifti1_extender
\brief This structure represents a 4-byte string that should follow the
binary nifti_1_header data in a NIFTI-1 header file. If the char
values are {1,0,0,0}, the file is expected to contain extensions,
values of {0,0,0,0} imply the file does not contain extensions.
Other sequences of values are not currently defined.
*/
struct nifti1_extender { char extension[4] ; } ;
typedef struct nifti1_extender nifti1_extender ;
/*! \struct nifti1_extension
\brief Data structure defining the fields of a header extension.
*/
struct nifti1_extension {
int esize ; /*!< size of extension, in bytes (must be multiple of 16) */
int ecode ; /*!< extension code, one of the NIFTI_ECODE_ values */
char * edata ; /*!< raw data, with no byte swapping (length is esize-8) */
} ;
typedef struct nifti1_extension nifti1_extension ;
/*---------------------------------------------------------------------------*/
/* DATA DIMENSIONALITY (as in ANALYZE 7.5):
---------------------------------------
dim[0] = number of dimensions;
- if dim[0] is outside range 1..7, then the header information
needs to be byte swapped appropriately
- ANALYZE supports dim[0] up to 7, but NIFTI-1 reserves
dimensions 1,2,3 for space (x,y,z), 4 for time (t), and
5,6,7 for anything else needed.
dim[i] = length of dimension #i, for i=1..dim[0] (must be positive)
- also see the discussion of intent_code, far below
pixdim[i] = voxel width along dimension #i, i=1..dim[0] (positive)
- cf. ORIENTATION section below for use of pixdim[0]
- the units of pixdim can be specified with the xyzt_units
field (also described far below).
Number of bits per voxel value is in bitpix, which MUST correspond with
the datatype field. The total number of bytes in the image data is
dim[1] * ... * dim[dim[0]] * bitpix / 8
In NIFTI-1 files, dimensions 1,2,3 are for space, dimension 4 is for time,
and dimension 5 is for storing multiple values at each spatiotemporal
voxel. Some examples:
- A typical whole-brain FMRI experiment's time series:
- dim[0] = 4
- dim[1] = 64 pixdim[1] = 3.75 xyzt_units = NIFTI_UNITS_MM
- dim[2] = 64 pixdim[2] = 3.75 | NIFTI_UNITS_SEC
- dim[3] = 20 pixdim[3] = 5.0
- dim[4] = 120 pixdim[4] = 2.0
- A typical T1-weighted anatomical volume:
- dim[0] = 3
- dim[1] = 256 pixdim[1] = 1.0 xyzt_units = NIFTI_UNITS_MM
- dim[2] = 256 pixdim[2] = 1.0
- dim[3] = 128 pixdim[3] = 1.1
- A single slice EPI time series:
- dim[0] = 4
- dim[1] = 64 pixdim[1] = 3.75 xyzt_units = NIFTI_UNITS_MM
- dim[2] = 64 pixdim[2] = 3.75 | NIFTI_UNITS_SEC
- dim[3] = 1 pixdim[3] = 5.0
- dim[4] = 1200 pixdim[4] = 0.2
- A 3-vector stored at each point in a 3D volume:
- dim[0] = 5
- dim[1] = 256 pixdim[1] = 1.0 xyzt_units = NIFTI_UNITS_MM
- dim[2] = 256 pixdim[2] = 1.0
- dim[3] = 128 pixdim[3] = 1.1
- dim[4] = 1 pixdim[4] = 0.0
- dim[5] = 3 intent_code = NIFTI_INTENT_VECTOR
- A single time series with a 3x3 matrix at each point:
- dim[0] = 5
- dim[1] = 1 xyzt_units = NIFTI_UNITS_SEC
- dim[2] = 1
- dim[3] = 1
- dim[4] = 1200 pixdim[4] = 0.2
- dim[5] = 9 intent_code = NIFTI_INTENT_GENMATRIX
- intent_p1 = intent_p2 = 3.0 (indicates matrix dimensions)
-----------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* DATA STORAGE:
------------
If the magic field is "n+1", then the voxel data is stored in the
same file as the header. In this case, the voxel data starts at offset
(int)vox_offset into the header file. Thus, vox_offset=352.0 means that
the data starts immediately after the NIFTI-1 header. If vox_offset is
greater than 352, the NIFTI-1 format does not say much about the
contents of the dataset file between the end of the header and the
start of the data.
FILES:
-----
If the magic field is "ni1", then the voxel data is stored in the
associated ".img" file, starting at offset 0 (i.e., vox_offset is not
used in this case, and should be set to 0.0).
When storing NIFTI-1 datasets in pairs of files, it is customary to name
the files in the pattern "name.hdr" and "name.img", as in ANALYZE 7.5.
When storing in a single file ("n+1"), the file name should be in
the form "name.nii" (the ".nft" and ".nif" suffixes are already taken;
cf. http://www.icdatamaster.com/n.html ).
BYTE ORDERING:
-------------
The byte order of the data arrays is presumed to be the same as the byte
order of the header (which is determined by examining dim[0]).
Floating point types are presumed to be stored in IEEE-754 format.
-----------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* DETAILS ABOUT vox_offset:
------------------------
In a .nii file, the vox_offset field value is interpreted as the start
location of the image data bytes in that file. In a .hdr/.img file pair,
the vox_offset field value is the start location of the image data
bytes in the .img file.
* If vox_offset is less than 352 in a .nii file, it is equivalent
to 352 (i.e., image data never starts before byte #352 in a .nii file).
* The default value for vox_offset in a .nii file is 352.
* In a .hdr file, the default value for vox_offset is 0.
* vox_offset should be an integer multiple of 16; otherwise, some
programs may not work properly (e.g., SPM). This is to allow
memory-mapped input to be properly byte-aligned.
Note that since vox_offset is an IEEE-754 32 bit float (for compatibility
with the ANALYZE-7.5 format), it effectively has a 24 bit mantissa. All
integers from 0 to 2^24 can be represented exactly in this format, but not
all larger integers are exactly storable as IEEE-754 32 bit floats. However,
unless you plan to have vox_offset be potentially larger than 16 MB, this
should not be an issue. (Actually, any integral multiple of 16 up to 2^27
can be represented exactly in this format, which allows for up to 128 MB
of random information before the image data. If that isn't enough, then
perhaps this format isn't right for you.)
In a .img file (i.e., image data stored separately from the NIfTI-1
header), data bytes between #0 and #vox_offset-1 (inclusive) are completely
undefined and unregulated by the NIfTI-1 standard. One potential use of
having vox_offset > 0 in the .hdr/.img file pair storage method is to make
the .img file be a copy of (or link to) a pre-existing image file in some
other format, such as DICOM; then vox_offset would be set to the offset of
the image data in this file. (It may not be possible to follow the
"multiple-of-16 rule" with an arbitrary external file; using the NIfTI-1
format in such a case may lead to a file that is incompatible with software
that relies on vox_offset being a multiple of 16.)
In a .nii file, data bytes between #348 and #vox_offset-1 (inclusive) may
be used to store user-defined extra information; similarly, in a .hdr file,
any data bytes after byte #347 are available for user-defined extra
information. The (very weak) regulation of this extra header data is
described elsewhere.
-----------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* DATA SCALING:
------------
If the scl_slope field is nonzero, then each voxel value in the dataset
should be scaled as
y = scl_slope * x + scl_inter
where x = voxel value stored
y = "true" voxel value
Normally, we would expect this scaling to be used to store "true" floating
values in a smaller integer datatype, but that is not required. That is,
it is legal to use scaling even if the datatype is a float type (crazy,
perhaps, but legal).
- However, the scaling is to be ignored if datatype is DT_RGB24.
- If datatype is a complex type, then the scaling is to be
applied to both the real and imaginary parts.
The cal_min and cal_max fields (if nonzero) are used for mapping (possibly
scaled) dataset values to display colors:
- Minimum display intensity (black) corresponds to dataset value cal_min.
- Maximum display intensity (white) corresponds to dataset value cal_max.
- Dataset values below cal_min should display as black also, and values
above cal_max as white.
- Colors "black" and "white", of course, may refer to any scalar display
scheme (e.g., a color lookup table specified via aux_file).
- cal_min and cal_max only make sense when applied to scalar-valued
datasets (i.e., dim[0] < 5 or dim[5] = 1).
-----------------------------------------------------------------------------*/
/*---------------------------------------------------------------------------*/
/* TYPE OF DATA (acceptable values for datatype field):
---------------------------------------------------
Values of datatype smaller than 256 are ANALYZE 7.5 compatible.
Larger values are NIFTI-1 additions. These are all multiples of 256, so
that no bits below position 8 are set in datatype. But there is no need
to use only powers-of-2, as the original ANALYZE 7.5 datatype codes do.
The additional codes are intended to include a complete list of basic
scalar types, including signed and unsigned integers from 8 to 64 bits,
floats from 32 to 128 bits, and complex (float pairs) from 64 to 256 bits.
Note that most programs will support only a few of these datatypes!
A NIFTI-1 program should fail gracefully (e.g., print a warning message)
when it encounters a dataset with a type it doesn't like.
-----------------------------------------------------------------------------*/
#undef DT_UNKNOWN /* defined in dirent.h on some Unix systems */
/*! \defgroup NIFTI1_DATATYPES
\brief nifti1 datatype codes
@{
*/
/*--- the original ANALYZE 7.5 type codes ---*/
#define DT_NONE 0
#define DT_UNKNOWN 0 /* what it says, dude */
#define DT_BINARY 1 /* binary (1 bit/voxel) */
#define DT_UNSIGNED_CHAR 2 /* unsigned char (8 bits/voxel) */
#define DT_SIGNED_SHORT 4 /* signed short (16 bits/voxel) */
#define DT_SIGNED_INT 8 /* signed int (32 bits/voxel) */
#define DT_FLOAT 16 /* float (32 bits/voxel) */
#define DT_COMPLEX 32 /* complex (64 bits/voxel) */
#define DT_DOUBLE 64 /* double (64 bits/voxel) */
#define DT_RGB 128 /* RGB triple (24 bits/voxel) */
#define DT_ALL 255 /* not very useful (?) */
/*----- another set of names for the same ---*/
#define DT_UINT8 2
#define DT_INT16 4
#define DT_INT32 8
#define DT_FLOAT32 16
#define DT_COMPLEX64 32
#define DT_FLOAT64 64
#define DT_RGB24 128
/*------------------- new codes for NIFTI ---*/
#define DT_INT8 256 /* signed char (8 bits) */
#define DT_UINT16 512 /* unsigned short (16 bits) */
#define DT_UINT32 768 /* unsigned int (32 bits) */
#define DT_INT64 1024 /* long long (64 bits) */
#define DT_UINT64 1280 /* unsigned long long (64 bits) */
#define DT_FLOAT128 1536 /* long double (128 bits) */
#define DT_COMPLEX128 1792 /* double pair (128 bits) */
#define DT_COMPLEX256 2048 /* long double pair (256 bits) */
#define DT_RGBA32 2304 /* 4 byte RGBA (32 bits/voxel) */
/* @} */
/*------- aliases for all the above codes ---*/
/*! \defgroup NIFTI1_DATATYPE_ALIASES
\brief aliases for the nifti1 datatype codes
@{
*/
/*! unsigned char. */
#define NIFTI_TYPE_UINT8 2
/*! signed short. */
#define NIFTI_TYPE_INT16 4
/*! signed int. */
#define NIFTI_TYPE_INT32 8
/*! 32 bit float. */
#define NIFTI_TYPE_FLOAT32 16
/*! 64 bit complex = 2 32 bit floats. */
#define NIFTI_TYPE_COMPLEX64 32
/*! 64 bit float = double. */
#define NIFTI_TYPE_FLOAT64 64
/*! 3 8 bit bytes. */
#define NIFTI_TYPE_RGB24 128
/*! signed char. */
#define NIFTI_TYPE_INT8 256
/*! unsigned short. */
#define NIFTI_TYPE_UINT16 512
/*! unsigned int. */
#define NIFTI_TYPE_UINT32 768
/*! signed long long. */
#define NIFTI_TYPE_INT64 1024
/*! unsigned long long. */
#define NIFTI_TYPE_UINT64 1280
/*! 128 bit float = long double. */
#define NIFTI_TYPE_FLOAT128 1536
/*! 128 bit complex = 2 64 bit floats. */
#define NIFTI_TYPE_COMPLEX128 1792
/*! 256 bit complex = 2 128 bit floats */
#define NIFTI_TYPE_COMPLEX256 2048
/*! 4 8 bit bytes. */
#define NIFTI_TYPE_RGBA32 2304
/* @} */
/*-------- sample typedefs for complicated types ---*/
#if 0
typedef struct { float r,i; } complex_float ;
typedef struct { double r,i; } complex_double ;
typedef struct { long double r,i; } complex_longdouble ;
typedef struct { unsigned char r,g,b; } rgb_byte ;
#endif
/*---------------------------------------------------------------------------*/
/* INTERPRETATION OF VOXEL DATA:
----------------------------
The intent_code field can be used to indicate that the voxel data has
some particular meaning. In particular, a large number of codes is
given to indicate that the the voxel data should be interpreted as
being drawn from a given probability distribution.
VECTOR-VALUED DATASETS:
----------------------
The 5th dimension of the dataset, if present (i.e., dim[0]=5 and
dim[5] > 1), contains multiple values (e.g., a vector) to be stored
at each spatiotemporal location. For example, the header values
- dim[0] = 5
- dim[1] = 64
- dim[2] = 64
- dim[3] = 20
- dim[4] = 1 (indicates no time axis)
- dim[5] = 3
- datatype = DT_FLOAT
- intent_code = NIFTI_INTENT_VECTOR
mean that this dataset should be interpreted as a 3D volume (64x64x20),
with a 3-vector of floats defined at each point in the 3D grid.
A program reading a dataset with a 5th dimension may want to reformat
the image data to store each voxels' set of values together in a struct
or array. This programming detail, however, is beyond the scope of the
NIFTI-1 file specification! Uses of dimensions 6 and 7 are also not
specified here.
STATISTICAL PARAMETRIC DATASETS (i.e., SPMs):
--------------------------------------------
Values of intent_code from NIFTI_FIRST_STATCODE to NIFTI_LAST_STATCODE
(inclusive) indicate that the numbers in the dataset should be interpreted
as being drawn from a given distribution. Most such distributions have
auxiliary parameters (e.g., NIFTI_INTENT_TTEST has 1 DOF parameter).
If the dataset DOES NOT have a 5th dimension, then the auxiliary parameters
are the same for each voxel, and are given in header fields intent_p1,
intent_p2, and intent_p3.
If the dataset DOES have a 5th dimension, then the auxiliary parameters
are different for each voxel. For example, the header values
- dim[0] = 5
- dim[1] = 128
- dim[2] = 128
- dim[3] = 1 (indicates a single slice)
- dim[4] = 1 (indicates no time axis)
- dim[5] = 2
- datatype = DT_FLOAT
- intent_code = NIFTI_INTENT_TTEST
mean that this is a 2D dataset (128x128) of t-statistics, with the
t-statistic being in the first "plane" of data and the degrees-of-freedom
parameter being in the second "plane" of data.
If the dataset 5th dimension is used to store the voxel-wise statistical
parameters, then dim[5] must be 1 plus the number of parameters required
by that distribution (e.g., intent_code=NIFTI_INTENT_TTEST implies dim[5]
must be 2, as in the example just above).
Note: intent_code values 2..10 are compatible with AFNI 1.5x (which is
why there is no code with value=1, which is obsolescent in AFNI).
OTHER INTENTIONS:
----------------
The purpose of the intent_* fields is to help interpret the values
stored in the dataset. Some non-statistical values for intent_code
and conventions are provided for storing other complex data types.
The intent_name field provides space for a 15 character (plus 0 byte)
'name' string for the type of data stored. Examples:
- intent_code = NIFTI_INTENT_ESTIMATE; intent_name = "T1";
could be used to signify that the voxel values are estimates of the
NMR parameter T1.
- intent_code = NIFTI_INTENT_TTEST; intent_name = "House";
could be used to signify that the voxel values are t-statistics
for the significance of 'activation' response to a House stimulus.
- intent_code = NIFTI_INTENT_DISPVECT; intent_name = "ToMNI152";
could be used to signify that the voxel values are a displacement
vector that transforms each voxel (x,y,z) location to the
corresponding location in the MNI152 standard brain.
- intent_code = NIFTI_INTENT_SYMMATRIX; intent_name = "DTI";
could be used to signify that the voxel values comprise a diffusion
tensor image.
If no data name is implied or needed, intent_name[0] should be set to 0.
-----------------------------------------------------------------------------*/
/*! default: no intention is indicated in the header. */
#define NIFTI_INTENT_NONE 0
/*-------- These codes are for probability distributions ---------------*/
/* Most distributions have a number of parameters,
below denoted by p1, p2, and p3, and stored in
- intent_p1, intent_p2, intent_p3 if dataset doesn't have 5th dimension
- image data array if dataset does have 5th dimension
Functions to compute with many of the distributions below can be found
in the CDF library from U Texas.
Formulas for and discussions of these distributions can be found in the
following books:
[U] Univariate Discrete Distributions,
NL Johnson, S Kotz, AW Kemp.
[C1] Continuous Univariate Distributions, vol. 1,
NL Johnson, S Kotz, N Balakrishnan.
[C2] Continuous Univariate Distributions, vol. 2,
NL Johnson, S Kotz, N Balakrishnan. */
/*----------------------------------------------------------------------*/
/*! [C2, chap 32] Correlation coefficient R (1 param):
p1 = degrees of freedom
R/sqrt(1-R*R) is t-distributed with p1 DOF. */
/*! \defgroup NIFTI1_INTENT_CODES
\brief nifti1 intent codes, to describe intended meaning of dataset contents
@{
*/
#define NIFTI_INTENT_CORREL 2
/*! [C2, chap 28] Student t statistic (1 param): p1 = DOF. */
#define NIFTI_INTENT_TTEST 3
/*! [C2, chap 27] Fisher F statistic (2 params):
p1 = numerator DOF, p2 = denominator DOF. */
#define NIFTI_INTENT_FTEST 4
/*! [C1, chap 13] Standard normal (0 params): Density = N(0,1). */
#define NIFTI_INTENT_ZSCORE 5
/*! [C1, chap 18] Chi-squared (1 param): p1 = DOF.
Density(x) proportional to exp(-x/2) * x^(p1/2-1). */
#define NIFTI_INTENT_CHISQ 6
/*! [C2, chap 25] Beta distribution (2 params): p1=a, p2=b.
Density(x) proportional to x^(a-1) * (1-x)^(b-1). */
#define NIFTI_INTENT_BETA 7
/*! [U, chap 3] Binomial distribution (2 params):
p1 = number of trials, p2 = probability per trial.
Prob(x) = (p1 choose x) * p2^x * (1-p2)^(p1-x), for x=0,1,...,p1. */
#define NIFTI_INTENT_BINOM 8
/*! [C1, chap 17] Gamma distribution (2 params):
p1 = shape, p2 = scale.
Density(x) proportional to x^(p1-1) * exp(-p2*x). */
#define NIFTI_INTENT_GAMMA 9
/*! [U, chap 4] Poisson distribution (1 param): p1 = mean.
Prob(x) = exp(-p1) * p1^x / x! , for x=0,1,2,.... */
#define NIFTI_INTENT_POISSON 10
/*! [C1, chap 13] Normal distribution (2 params):
p1 = mean, p2 = standard deviation. */
#define NIFTI_INTENT_NORMAL 11
/*! [C2, chap 30] Noncentral F statistic (3 params):
p1 = numerator DOF, p2 = denominator DOF,
p3 = numerator noncentrality parameter. */
#define NIFTI_INTENT_FTEST_NONC 12
/*! [C2, chap 29] Noncentral chi-squared statistic (2 params):
p1 = DOF, p2 = noncentrality parameter. */
#define NIFTI_INTENT_CHISQ_NONC 13
/*! [C2, chap 23] Logistic distribution (2 params):
p1 = location, p2 = scale.
Density(x) proportional to sech^2((x-p1)/(2*p2)). */
#define NIFTI_INTENT_LOGISTIC 14
/*! [C2, chap 24] Laplace distribution (2 params):
p1 = location, p2 = scale.
Density(x) proportional to exp(-abs(x-p1)/p2). */
#define NIFTI_INTENT_LAPLACE 15
/*! [C2, chap 26] Uniform distribution: p1 = lower end, p2 = upper end. */
#define NIFTI_INTENT_UNIFORM 16
/*! [C2, chap 31] Noncentral t statistic (2 params):
p1 = DOF, p2 = noncentrality parameter. */
#define NIFTI_INTENT_TTEST_NONC 17
/*! [C1, chap 21] Weibull distribution (3 params):
p1 = location, p2 = scale, p3 = power.
Density(x) proportional to
((x-p1)/p2)^(p3-1) * exp(-((x-p1)/p2)^p3) for x > p1. */
#define NIFTI_INTENT_WEIBULL 18
/*! [C1, chap 18] Chi distribution (1 param): p1 = DOF.
Density(x) proportional to x^(p1-1) * exp(-x^2/2) for x > 0.
p1 = 1 = 'half normal' distribution
p1 = 2 = Rayleigh distribution
p1 = 3 = Maxwell-Boltzmann distribution. */
#define NIFTI_INTENT_CHI 19
/*! [C1, chap 15] Inverse Gaussian (2 params):
p1 = mu, p2 = lambda
Density(x) proportional to
exp(-p2*(x-p1)^2/(2*p1^2*x)) / x^3 for x > 0. */
#define NIFTI_INTENT_INVGAUSS 20
/*! [C2, chap 22] Extreme value type I (2 params):
p1 = location, p2 = scale
cdf(x) = exp(-exp(-(x-p1)/p2)). */
#define NIFTI_INTENT_EXTVAL 21
/*! Data is a 'p-value' (no params). */
#define NIFTI_INTENT_PVAL 22
/*! Data is ln(p-value) (no params).
To be safe, a program should compute p = exp(-abs(this_value)).
The nifti_stats.c library returns this_value
as positive, so that this_value = -log(p). */
#define NIFTI_INTENT_LOGPVAL 23
/*! Data is log10(p-value) (no params).
To be safe, a program should compute p = pow(10.,-abs(this_value)).
The nifti_stats.c library returns this_value
as positive, so that this_value = -log10(p). */
#define NIFTI_INTENT_LOG10PVAL 24
/*! Smallest intent_code that indicates a statistic. */
#define NIFTI_FIRST_STATCODE 2
/*! Largest intent_code that indicates a statistic. */
#define NIFTI_LAST_STATCODE 24
/*---------- these values for intent_code aren't for statistics ----------*/
/*! To signify that the value at each voxel is an estimate
of some parameter, set intent_code = NIFTI_INTENT_ESTIMATE.
The name of the parameter may be stored in intent_name. */
#define NIFTI_INTENT_ESTIMATE 1001
/*! To signify that the value at each voxel is an index into
some set of labels, set intent_code = NIFTI_INTENT_LABEL.
The filename with the labels may stored in aux_file. */
#define NIFTI_INTENT_LABEL 1002
/*! To signify that the value at each voxel is an index into the
NeuroNames labels set, set intent_code = NIFTI_INTENT_NEURONAME. */
#define NIFTI_INTENT_NEURONAME 1003
/*! To store an M x N matrix at each voxel:
- dataset must have a 5th dimension (dim[0]=5 and dim[5]>1)
- intent_code must be NIFTI_INTENT_GENMATRIX
- dim[5] must be M*N
- intent_p1 must be M (in float format)
- intent_p2 must be N (ditto)
- the matrix values A[i][[j] are stored in row-order:
- A[0][0] A[0][1] ... A[0][N-1]
- A[1][0] A[1][1] ... A[1][N-1]
- etc., until
- A[M-1][0] A[M-1][1] ... A[M-1][N-1] */
#define NIFTI_INTENT_GENMATRIX 1004
/*! To store an NxN symmetric matrix at each voxel:
- dataset must have a 5th dimension
- intent_code must be NIFTI_INTENT_SYMMATRIX
- dim[5] must be N*(N+1)/2
- intent_p1 must be N (in float format)
- the matrix values A[i][[j] are stored in row-order:
- A[0][0]
- A[1][0] A[1][1]
- A[2][0] A[2][1] A[2][2]
- etc.: row-by-row */
#define NIFTI_INTENT_SYMMATRIX 1005
/*! To signify that the vector value at each voxel is to be taken
as a displacement field or vector:
- dataset must have a 5th dimension
- intent_code must be NIFTI_INTENT_DISPVECT
- dim[5] must be the dimensionality of the displacment
vector (e.g., 3 for spatial displacement, 2 for in-plane) */
#define NIFTI_INTENT_DISPVECT 1006 /* specifically for displacements */
#define NIFTI_INTENT_VECTOR 1007 /* for any other type of vector */
/*! To signify that the vector value at each voxel is really a
spatial coordinate (e.g., the vertices or nodes of a surface mesh):
- dataset must have a 5th dimension
- intent_code must be NIFTI_INTENT_POINTSET
- dim[0] = 5
- dim[1] = number of points
- dim[2] = dim[3] = dim[4] = 1
- dim[5] must be the dimensionality of space (e.g., 3 => 3D space).
- intent_name may describe the object these points come from
(e.g., "pial", "gray/white" , "EEG", "MEG"). */
#define NIFTI_INTENT_POINTSET 1008
/*! To signify that the vector value at each voxel is really a triple
of indexes (e.g., forming a triangle) from a pointset dataset:
- dataset must have a 5th dimension
- intent_code must be NIFTI_INTENT_TRIANGLE
- dim[0] = 5
- dim[1] = number of triangles
- dim[2] = dim[3] = dim[4] = 1
- dim[5] = 3
- datatype should be an integer type (preferably DT_INT32)
- the data values are indexes (0,1,...) into a pointset dataset. */
#define NIFTI_INTENT_TRIANGLE 1009
/*! To signify that the vector value at each voxel is a quaternion:
- dataset must have a 5th dimension
- intent_code must be NIFTI_INTENT_QUATERNION
- dim[0] = 5
- dim[5] = 4
- datatype should be a floating point type */
#define NIFTI_INTENT_QUATERNION 1010
/*! Dimensionless value - no params - although, as in _ESTIMATE
the name of the parameter may be stored in intent_name. */
#define NIFTI_INTENT_DIMLESS 1011
/*---------- these values apply to GIFTI datasets ----------*/
/*! To signify that the value at each location is from a time series. */
#define NIFTI_INTENT_TIME_SERIES 2001
/*! To signify that the value at each location is a node index, from
a complete surface dataset. */
#define NIFTI_INTENT_NODE_INDEX 2002
/*! To signify that the vector value at each location is an RGB triplet,
of whatever type.
- dataset must have a 5th dimension
- dim[0] = 5
- dim[1] = number of nodes
- dim[2] = dim[3] = dim[4] = 1
- dim[5] = 3
*/
#define NIFTI_INTENT_RGB_VECTOR 2003
/*! To signify that the vector value at each location is a 4 valued RGBA
vector, of whatever type.
- dataset must have a 5th dimension
- dim[0] = 5
- dim[1] = number of nodes
- dim[2] = dim[3] = dim[4] = 1
- dim[5] = 4
*/
#define NIFTI_INTENT_RGBA_VECTOR 2004
/*! To signify that the value at each location is a shape value, such
as the curvature. */
#define NIFTI_INTENT_SHAPE 2005
/* @} */
/*---------------------------------------------------------------------------*/
/* 3D IMAGE (VOLUME) ORIENTATION AND LOCATION IN SPACE:
---------------------------------------------------
There are 3 different methods by which continuous coordinates can
attached to voxels. The discussion below emphasizes 3D volumes, and
the continuous coordinates are referred to as (x,y,z). The voxel
index coordinates (i.e., the array indexes) are referred to as (i,j,k),
with valid ranges:
i = 0 .. dim[1]-1
j = 0 .. dim[2]-1 (if dim[0] >= 2)
k = 0 .. dim[3]-1 (if dim[0] >= 3)
The (x,y,z) coordinates refer to the CENTER of a voxel. In methods
2 and 3, the (x,y,z) axes refer to a subject-based coordinate system,
with
+x = Right +y = Anterior +z = Superior.
This is a right-handed coordinate system. However, the exact direction
these axes point with respect to the subject depends on qform_code
(Method 2) and sform_code (Method 3).
N.B.: The i index varies most rapidly, j index next, k index slowest.
Thus, voxel (i,j,k) is stored starting at location
(i + j*dim[1] + k*dim[1]*dim[2]) * (bitpix/8)
into the dataset array.
N.B.: The ANALYZE 7.5 coordinate system is
+x = Left +y = Anterior +z = Superior
which is a left-handed coordinate system. This backwardness is
too difficult to tolerate, so this NIFTI-1 standard specifies the
coordinate order which is most common in functional neuroimaging.
N.B.: The 3 methods below all give the locations of the voxel centers
in the (x,y,z) coordinate system. In many cases, programs will wish
to display image data on some other grid. In such a case, the program
will need to convert its desired (x,y,z) values into (i,j,k) values
in order to extract (or interpolate) the image data. This operation
would be done with the inverse transformation to those described below.
N.B.: Method 2 uses a factor 'qfac' which is either -1 or 1; qfac is
stored in the otherwise unused pixdim[0]. If pixdim[0]=0.0 (which
should not occur), we take qfac=1. Of course, pixdim[0] is only used
when reading a NIFTI-1 header, not when reading an ANALYZE 7.5 header.
N.B.: The units of (x,y,z) can be specified using the xyzt_units field.