Sample compressed JPEG/JPEG-LS/JPEG 2000 images revised
David Clunie
Dave Heaney pointed out that some of the fragments in the
compressed data set samples had odd lengths, which is
not allowed. The entire image set has been revised
accordingly. See:
ftp://medical.nema.org/MEDICAL/Dicom/DataSets/WG04
david
PS. The README file is included below ...
2003/12/06
This set consists of a collection of medical images in both
uncompressed form, and compressed with various lossless
and lossy compression schemes defined by the DICOM standard.
These include JPEG lossless (JPLL) and lossy (JPLY), JPEG-LS in
lossless (JLSL) and near-lossless modes (JLSN), and JPEG 2000
reversible (J2KR) and irreversible (J2KI).
Not all the images have been compressed using all schemes and
modes; in particular the color images have been compressed using
JPEG 2000 only.
No effort has been made to align the target bit-rates (compression
ratios) of the various lossy compressors, since the objective
is to provide sample images to test and demonstrate implementation,
not to compare the efficacy of one scheme with another. While one
can obviously compare the effectiveness of lossless compression
by looking at the file sizes, the small number of images and diversity
of modalities used make such a comparison fairly meaningless. For
interest, the file sizes are as follows:
IMG UNC J2KI J2KR JLSL JLSN JPLL JPLY
CT1 530632 13796 180658 170640 91332 210288 67244
CT2 525920 6468 121098 117046 50910 165882 53934
MG1 28582296 2680530 12233088 12022600 6572812 12833974 2823062
MR1 525926 8652 238498 231334 128648 261890 71610
MR2 2098988 113192 588956 599750 281374 779466 120052
MR3 533746 12068 122016 127538 73606 172860 46348
MR4 526146 12036 114644 118548 47530 155168 17454
NM1 527066 2944 89700 91778 31880 118742 9504
RG1 7200056 676486 4284936 4197418 2765246 4445380 1876916
RG2 7534130 707292 1660120 1672750 698940 2162250 211226
RG3 6196600 206790 831844 848764 289688 1398600 93742
SC1 10187866 781316 2443898 2153920 1102216 3111760 768020
US1 922878 58776 153498
VL1 1103386 69880 296898
VL2 1103384 69918 308566
VL3 1103396 69874 304014
VL4 12475634 780748 4159712
VL5 26754534 1673328 8919932
VL6 1103396 69650 368514
XA1 2098206 108954 403052 391664 131446 495356 43800
The compressed images are encapsulated as per the standard
requirements in PS 3.5; even though all the images are single
frame in some cases the compressed bitstream is fragmented
into multiple items.
The codecs used were:
- for JPEG, the Stanford PVRG codec
- for JPEG-LS, the dicom3tools rawnjl2 codec
- for JPEG 2000, the Kakadu codec
The following images are contained in the set, as described
by the accompanying DICOMDIR:
PATIENT CompressedSamples^XA1 20XA1
STUDY 20XA1 20030612 185520
SERIES 1 XA
IMAGE 1
-> IMAGES\REF\XA1_UNC
IMAGE 7
-> IMAGES\JLSN\XA1_JLSN
IMAGE 6
-> IMAGES\JLSL\XA1_JLSL
IMAGE 5
-> IMAGES\JPLY\XA1_JPLY
IMAGE 4
-> IMAGES\JPLL\XA1_JPLL
IMAGE 2
-> IMAGES\J2KR\XA1_J2KR
IMAGE 3
-> IMAGES\J2KI\XA1_J2KI
PATIENT CompressedSamples^VL6 19VL6
STUDY 19VL6 20030612 185520 Stomach Body Biopsy - Micro
SERIES 1 OT
IMAGE 1
-> IMAGES\REF\VL6_UNC
IMAGE 2
-> IMAGES\J2KR\VL6_J2KR
IMAGE 3
-> IMAGES\J2KI\VL6_J2KI
PATIENT CompressedSamples^VL5 18VL5
STUDY 18VL5 20030612 185520 Whole Prostate
SERIES 1 OT
IMAGE 1
-> IMAGES\REF\VL5_UNC
IMAGE 2
-> IMAGES\J2KR\VL5_J2KR
IMAGE 3
-> IMAGES\J2KI\VL5_J2KI
PATIENT CompressedSamples^VL4 17VL4
STUDY 17VL4 20030612 185520 Fundoscopy
SERIES 1 OT
IMAGE 1
-> IMAGES\REF\VL4_UNC
IMAGE 2
-> IMAGES\J2KR\VL4_J2KR
IMAGE 3
-> IMAGES\J2KI\VL4_J2KI
PATIENT CompressedSamples^VL3 16VL3
STUDY 16VL3 20030612 185520 Arterio-venous Malformation
SERIES 1 OT
IMAGE 1
-> IMAGES\REF\VL3_UNC
IMAGE 2
-> IMAGES\J2KR\VL3_J2KR
IMAGE 3
-> IMAGES\J2KI\VL3_J2KI
PATIENT CompressedSamples^VL2 15VL2
STUDY 15VL2 20030612 185520 Prepyloric Ulcer
SERIES 1 OT
IMAGE 1
-> IMAGES\REF\VL2_UNC
IMAGE 2
-> IMAGES\J2KR\VL2_J2KR
IMAGE 3
-> IMAGES\J2KI\VL2_J2KI
PATIENT CompressedSamples^VL1 14VL1
STUDY 14VL1 20030612 185520 Barretts Esophagus
SERIES 1 OT
IMAGE 1
-> IMAGES\REF\VL1_UNC
IMAGE 2
-> IMAGES\J2KR\VL1_J2KR
IMAGE 3
-> IMAGES\J2KI\VL1_J2KI
PATIENT CompressedSamples^US1 13US1
STUDY 13US1 20030612 185520
SERIES 1 US
IMAGE 1
-> IMAGES\REF\US1_UNC
IMAGE 2
-> IMAGES\J2KR\US1_J2KR
IMAGE 3
-> IMAGES\J2KI\US1_J2KI
PATIENT CompressedSamples^SC1 12SC1
STUDY 12SC1 20030612 185520
SERIES 1 CT
IMAGE 1
-> IMAGES\REF\SC1_UNC
IMAGE 7
-> IMAGES\JLSN\SC1_JLSN
IMAGE 6
-> IMAGES\JLSL\SC1_JLSL
IMAGE 5
-> IMAGES\JPLY\SC1_JPLY
IMAGE 4
-> IMAGES\JPLL\SC1_JPLL
IMAGE 2
-> IMAGES\J2KR\SC1_J2KR
IMAGE 3
-> IMAGES\J2KI\SC1_J2KI
PATIENT CompressedSamples^RG3 11RG3
STUDY 11RG3 FUJI95706 20030612 185520 Non-ossifying fibroma of distal tibia
SERIES 1 CR
IMAGE 1
-> IMAGES\REF\RG3_UNC
IMAGE 7
-> IMAGES\JLSN\RG3_JLSN
IMAGE 6
-> IMAGES\JLSL\RG3_JLSL
IMAGE 5
-> IMAGES\JPLY\RG3_JPLY
IMAGE 4
-> IMAGES\JPLL\RG3_JPLL
IMAGE 2
-> IMAGES\J2KR\RG3_J2KR
IMAGE 3
-> IMAGES\J2KI\RG3_J2KI
PATIENT CompressedSamples^RG2 10RG2
STUDY 10RG2 20030612 185520 HIP JOINT
SERIES 1 CR
IMAGE 1
-> IMAGES\REF\RG2_UNC
IMAGE 7
-> IMAGES\JLSN\RG2_JLSN
IMAGE 6
-> IMAGES\JLSL\RG2_JLSL
IMAGE 5
-> IMAGES\JPLY\RG2_JPLY
IMAGE 4
-> IMAGES\JPLL\RG2_JPLL
IMAGE 2
-> IMAGES\J2KR\RG2_J2KR
IMAGE 3
-> IMAGES\J2KI\RG2_J2KI
PATIENT CompressedSamples^RG1 9RG1
STUDY 9RG1 exam Number 20030612 185520 THORAX
SERIES 1 CR
IMAGE 1
-> IMAGES\REF\RG1_UNC
IMAGE 7
-> IMAGES\JLSN\RG1_JLSN
IMAGE 6
-> IMAGES\JLSL\RG1_JLSL
IMAGE 5
-> IMAGES\JPLY\RG1_JPLY
IMAGE 4
-> IMAGES\JPLL\RG1_JPLL
IMAGE 2
-> IMAGES\J2KR\RG1_J2KR
IMAGE 3
-> IMAGES\J2KI\RG1_J2KI
PATIENT CompressedSamples^NM1 8NM1
STUDY 8NM1 20030612 185520 Whole Body Bone
SERIES 1 NM
IMAGE 1
-> IMAGES\REF\NM1_UNC
IMAGE 7
-> IMAGES\JLSN\NM1_JLSN
IMAGE 6
-> IMAGES\JLSL\NM1_JLSL
IMAGE 5
-> IMAGES\JPLY\NM1_JPLY
IMAGE 4
-> IMAGES\JPLL\NM1_JPLL
IMAGE 2
-> IMAGES\J2KR\NM1_J2KR
IMAGE 3
-> IMAGES\J2KI\NM1_J2KI
PATIENT CompressedSamples^MR4 7MR4
STUDY 7MR4 20030612 185520 BRAIN
SERIES 1 MR
IMAGE 1
-> IMAGES\REF\MR4_UNC
IMAGE 7
-> IMAGES\JLSN\MR4_JLSN
IMAGE 6
-> IMAGES\JLSL\MR4_JLSL
IMAGE 5
-> IMAGES\JPLY\MR4_JPLY
IMAGE 4
-> IMAGES\JPLL\MR4_JPLL
IMAGE 2
-> IMAGES\J2KR\MR4_J2KR
IMAGE 3
-> IMAGES\J2KI\MR4_J2KI
PATIENT CompressedSamples^MR3 6MR3
STUDY 6MR3 20030612 185520 e+1 KNEE-RT.
SERIES 1 MR
IMAGE 1
-> IMAGES\REF\MR3_UNC
IMAGE 7
-> IMAGES\JLSN\MR3_JLSN
IMAGE 6
-> IMAGES\JLSL\MR3_JLSL
IMAGE 5
-> IMAGES\JPLY\MR3_JPLY
IMAGE 4
-> IMAGES\JPLL\MR3_JPLL
IMAGE 2
-> IMAGES\J2KR\MR3_J2KR
IMAGE 3
-> IMAGES\J2KI\MR3_J2KI
PATIENT CompressedSamples^MR2 5MR2
STUDY 5MR2 20030612 185520 SHOULDER
SERIES 1 MR
IMAGE 1
-> IMAGES\REF\MR2_UNC
IMAGE 7
-> IMAGES\JLSN\MR2_JLSN
IMAGE 6
-> IMAGES\JLSL\MR2_JLSL
IMAGE 5
-> IMAGES\JPLY\MR2_JPLY
IMAGE 4
-> IMAGES\JPLL\MR2_JPLL
IMAGE 2
-> IMAGES\J2KR\MR2_J2KR
IMAGE 3
-> IMAGES\J2KI\MR2_J2KI
PATIENT CompressedSamples^MR1 4MR1
STUDY 4MR1 20030612 185520
SERIES 1 MR
IMAGE 1
-> IMAGES\REF\MR1_UNC
IMAGE 7
-> IMAGES\JLSN\MR1_JLSN
IMAGE 6
-> IMAGES\JLSL\MR1_JLSL
IMAGE 5
-> IMAGES\JPLY\MR1_JPLY
IMAGE 4
-> IMAGES\JPLL\MR1_JPLL
IMAGE 2
-> IMAGES\J2KR\MR1_J2KR
IMAGE 3
-> IMAGES\J2KI\MR1_J2KI
PATIENT CompressedSamples^MG1 3MG1
STUDY 3MG1 20030612 185520 Malignant mass with micro-calcifications
SERIES 1 RG
IMAGE 1
-> IMAGES\REF\MG1_UNC
IMAGE 7
-> IMAGES\JLSN\MG1_JLSN
IMAGE 6
-> IMAGES\JLSL\MG1_JLSL
IMAGE 5
-> IMAGES\JPLY\MG1_JPLY
IMAGE 4
-> IMAGES\JPLL\MG1_JPLL
IMAGE 2
-> IMAGES\J2KR\MG1_J2KR
IMAGE 3
-> IMAGES\J2KI\MG1_J2KI
PATIENT CompressedSamples^CT2 2CT2
STUDY 2CT2 20030612 185520
SERIES 1 CT
IMAGE 1
-> IMAGES\REF\CT2_UNC
IMAGE 7
-> IMAGES\JLSN\CT2_JLSN
IMAGE 6
-> IMAGES\JLSL\CT2_JLSL
IMAGE 5
-> IMAGES\JPLY\CT2_JPLY
IMAGE 4
-> IMAGES\JPLL\CT2_JPLL
IMAGE 2
-> IMAGES\J2KR\CT2_J2KR
IMAGE 3
-> IMAGES\J2KI\CT2_J2KI
PATIENT CompressedSamples^CT1 1CT1
STUDY 1CT1 20030612 185520 e+1
SERIES 1 CT
IMAGE 1
-> IMAGES\REF\CT1_UNC
IMAGE 7
-> IMAGES\JLSN\CT1_JLSN
IMAGE 6
-> IMAGES\JLSL\CT1_JLSL
IMAGE 5
-> IMAGES\JPLY\CT1_JPLY
IMAGE 4
-> IMAGES\JPLL\CT1_JPLL
IMAGE 2
-> IMAGES\J2KR\CT1_J2KR
IMAGE 3
-> IMAGES\J2KI\CT1_J2KI
The set of uncompressed images are those that were originally
supplied to ISO/IEC JTC1/SC29/WG1 for the purpose of informally
evaluating the effect of proposed JPEG 2000 schemes on image
quality. An updated version of the original description of
the images supplied to JTC1/SC29/WG1 follows ...
The images contain no patient identifying information.
They were initially obtained from various publically available
collections of images, including NEMA DICOM demonstrations
and public web servers, as well as some images specifically
solicited for the purpose of testing compression. The original
sources were:
- CT1,MR3,NM1,US1,XA1: GE
- CT2,MR1: Toshiba
- MR2,MR4,RG1: Philips
- RG2,RG3: Fuji
- MG1: USF (http://marathon.csee.usf.edu/Mammography/Database.html)
- SC1: Cemax (now Kodak)
- VL1,VL2,VL3: ASGE
- VL4: AAO (Hildebrand)
- VL5,VL6: CAP (Balis)
They are intended to be a representative set of the broad
range of different types of medical images. They may not
necessarily be particularly challenging or unusual and may not
contain any particularly subtle abnormalities, so should
not be assumed to be critical tests of a compression
scheme nor imply any likely usefulness from a diagnostic
point of view. The decision to use a lossy compression
scheme for medical applications may require a more rigorous
observer-performance study of a carefully selected set to
test a particular diagnostic task.
The uncompressed images are "dual-personality" in the sense that
they are both DICOM and TIFF images.
Images are all encoded as standard DICOM Image Storage SOP Classes;
the Visible Light and Mammography images are stored as Secondary Capture,
since more specific SOP Classes were not defined at the time that
these images were originally created.
Non-medical users may find it more convenient to extract
the pixel data from the images using a conventional TIFF
reader, rather than a DICOM reader. Be aware that some of
the images, including the CT and some of the MR images,
contained signed data, and hence will "look strange". It
is expected that the results of the compress/decompress
cycle will result in data of the same form as the original
and will be evaluated visually using medical software
that "knows" about signed data, ie. don't try to fix
what these "look like" before compressing them !
The following uncompressed reference images are present:
Image File Pixel Data Rows Cols Bits Samples Zero Order
Bytes Offset Entropy
CT Lung Contrast CT1_UNC 524288 6176 512 512 16 1 8.20914
CT Brain Contrast CT2_UNC 524288 1426 512 512 16 1 6.8012
Mammogram MG1_UNC 28580992 1186 4664 3064 12 1 10.561
MR Posterior Fossa MR1_UNC 524288 1430 512 512 16 1 10.083
MR Shoulder MR2_UNC 2097152 1728 1024 1024 12 1 6.89806
MR Knee MR3_UNC 524288 9272 512 512 16 1 6.30807
MR Brain MR4_UNC 524288 1748 512 512 12 1 6.03216
NM Bonescan NM1_UNC 524288 2632 1024 256 16 1 3.99372
Chest X-Ray RG1_UNC 7198310 1570 1955 1841 15 1 13.7073
Hip Joint RG2_UNC 7532800 1134 2140 1760 10 1 7.20069
Bone Lesion Leg RG3_UNC 6195200 1202 1760 1760 10 1 6.62469
US Color Tissue Doppler US1_UNC 921600 1044 480 640 8 3 3.11635
XA Neuroangiogram XA1_UNC 2097152 936 1024 1024 10 1 6.53488
Scanned Film CT Brain SC1_UNC 10186752 880 2487 2048 12 1 5.75373
GI Endoscopy Esophagus VL1_UNC 1102248 894 486 756 8 3 2.66609
GI Endoscopy Stomach VL2_UNC 1102248 890 486 756 8 3 2.99886
GI Endoscopy Colon VL3_UNC 1102248 900 486 756 8 3 2.55583
Retinal Fundoscopy VL4_UNC 12474504 870 1868 2226 8 3 4.32233
Pathology Whole Prostate VL5_UNC 26753400 886 3340 2670 8 3 6.16013
Pathology Stomach Biopsy VL6_UNC 1102248 902 486 756 8 3 4.53285
Note that Bits is the official number of bits used as
stated by the source of the images ... the dynamic
range of the actual image may not use this much.
Also, the images are stored as 8 or 16 bit words (little
endian), eg. a 12 bit pixel actually occupies 16 bits (in
the least significant bits). This is all described in the
DICOM header, but the TIFF header doesn't preserve this
distinction.
All images are single frame. The color image(s) are all
color-by pixel (ie. RGBRGB...), not color-by-plane.
CHANGES.
2003/12/06
- make fragment items padded to even length if necessary
- fragment more aggressively to better test decoders
- revise file sizes in README
2003/07/10
- initial release.
David Clunie
The data set has been updated again to account for some
lossy JPEG bugs that Marco Eichelberg found (wrong
transfer syntax for > 8 bit images and precision not always
12 for lossy JPEG images), as well as to more properly
populate the descriptive attributes with DERIVED image type,
descriptions and coded sequences for derivation as well as
source image references.
Thanks to all those folks who are reviewing these images and
giving regular feedback.
In addition, a new tar file has been included that contains
RLE compressed images produced by Marco Eichelberg. See:
ftp://medical.nema.org/MEDICAL/Dicom/DataSets/WG04/compsamples_rle.tar.bz2
david
David Clunie <dcl...@dclunie.com> wrote in message news:<3FD3249E...@dclunie.com>...