Fail to load pet/ct Gemini

Support
1

Operating system: Mac OS X
Slicer version:4.10.2
Expected behavior: load pet/ct acquired by Phillips Gemini
Actual behavior: It returns to “Could not load: 7: bone as a Scalar Volume”, the metadata is as following. None of the files for one patient can be opened, pet or CTs. Do you know what’s the problem? Data in https://www.dropbox.com/sh/m0e7t4rahk2ykns/AABLbj8nlGAWtGdNowv1BTNwa?dl=0

Thanks
Shuang

[0008,0005] SpecificCharacterSet ISO_IR 100 CS 10
[0008,0008] ImageType [3] ORIGINAL, PRIMARY, AXIAL CS 22
[0008,0012] InstanceCreationDate 20190628 DA 8
[0008,0013] InstanceCreationTime 101200 TM 6
[0008,0016] SOPClassUID 1.2.840.10008.5.1.4.1.1.2 UI 26
[0008,0018] SOPInstanceUID 1.2.840.113704.1.111.8320.1561687920.11509 UI 42
[0008,0020] StudyDate 20190628 DA 8
[0008,0022] AcquisitionDate 20190628 DA 8
[0008,0023] ContentDate 20190628 DA 8
[0008,0030] StudyTime 095513 TM 6
[0008,0032] AcquisitionTime 101052 TM 6
[0008,0033] ContentTime 101108.701 TM 10
[0008,0050] AccessionNumber 1906289110 SH 10
[0008,0060] Modality CT CS 2
[0008,0070] Manufacturer Philips LO 8
[0008,0080] InstitutionName PMSTL LO 6
[0008,0081] InstitutionAddress Haifa, MATAM ST 12
[0008,0090] ReferringPhysicianName PET-11909-2 PN 12
[0008,1010] StationName PHILIPS-0CE7C19 SH 16
[0008,1030] StudyDescription Body LO 4
[0008,103e] SeriesDescription bone LO 4
[0008,1040] InstitutionalDepartmentName Radiology LO 10
[0008,1070] OperatorsName PN 0
[0008,1090] ManufacturerModelName GEMINI TF TOF 16 LO 16
[0008,1110] ReferencedStudySequence SQ 66
[fffe,e000] Item na 50
[0008,1150] ReferencedSOPClassUID 1.2.840.10008.3.1.2.3.1 UI 24
[0008,1155] ReferencedSOPInstanceUID 1906289110 UI 10
[0008,1111] ReferencedPerformedProcedureStepSequence SQ 96
[fffe,e000] Item na 80
[0008,1150] ReferencedSOPClassUID 1.2.840.10008.3.1.2.3.3 UI 24
[0008,1155] ReferencedSOPInstanceUID 1.2.840.113704.1.111.2768.1561686900.44 UI 40
[0008,1120] ReferencedPatientSequence SQ 32
[fffe,e000] Item na 16
[0008,1150] ReferencedSOPClassUID UI 0
[0008,1155] ReferencedSOPInstanceUID UI 0
[0008,1140] ReferencedImageSequence SQ 100
[fffe,e000] Item na 84
[0008,1150] ReferencedSOPClassUID 1.2.840.10008.5.1.4.1.1.2 UI 26
[0008,1155] ReferencedSOPInstanceUID 1.2.840.113704.1.111.8320.1561686995.11013 UI 42
[0010,0010] PatientName DIAO SHU ZHEN PN 14
[0010,0020] PatientID PET-11909-2 LO 12
[0010,0030] PatientBirthDate 19430628 DA 8
[0010,0040] PatientSex F CS 2
[0010,1000] RETIRED_OtherPatientIDs LO 0
[0010,1010] PatientAge 076Y AS 4
[0010,1020] PatientSize 1.53 DS 4
[0010,1030] PatientWeight 58 DS 2
[0010,2000] MedicalAlerts LO 0
[0010,2110] Allergies LO 0
[0010,2160] EthnicGroup SH 0
[0010,21b0] AdditionalPatientHistory LT 0
[0010,21c0] PregnancyStatus US 0
[0010,4000] PatientComments LT 0
[0018,0022] ScanOptions AXIAL CS 6
[0018,0050] SliceThickness 3 DS 2
[0018,0060] KVP 120 DS 4
[0018,0090] DataCollectionDiameter 500 DS 4
[0018,1020] SoftwareVersions 2.3.0 LO 6
[0018,1030] ProtocolName Brain C-/Head LO 14
[0018,1100] ReconstructionDiameter 250 DS 4
[0018,1120] GantryDetectorTilt 0 DS 2
[0018,1130] TableHeight 185 DS 4
[0018,1140] RotationDirection CW CS 2
[0018,1143] ScanArc 360 DS 4
[0018,1150] ExposureTime 1500 IS 4
[0018,1151] XRayTubeCurrent 333 IS 4
[0018,1152] Exposure 500 IS 4
[0018,1160] FilterType D SH 2
[0018,1210] ConvolutionKernel D SH 2
[0018,5100] PatientPosition HFS CS 4
[0018,9321] CTExposureSequence SQ 96
[fffe,e000] Item na 80
[0018,9324] EstimatedDoseSaving 0 FD 8
[0018,9328] ExposureTimeInms 1.5015015015015014 FD 8
[0018,9330] XRayTubeCurrentInmA 333 FD 8
[0018,9332] ExposureInmAs 500 FD 8
[0018,9345] CTDIvol 69.200000000000003 FD 8
[0018,9323] ExposureModulationType CS 0
[0018,9345] CTDIvol 69.200000000000003 FD 8
[0020,000d] StudyInstanceUID 1906289110 UI 10
[0020,000e] SeriesInstanceUID 1.2.840.113704.1.111.2744.1561687837.23 UI 40
[0020,0010] StudyID 50431 SH 6
[0020,0011] SeriesNumber 7 IS 2
[0020,0012] AcquisitionNumber IS 0
[0020,0013] InstanceNumber 41 IS 2
[0020,0032] ImagePositionPatient [3] -141, -55, 1781.5 DS 16
[0020,0037] ImageOrientationPatient [6] 1, 0, 0, 0, 1, 0 DS 12
[0020,0052] FrameOfReferenceUID 1.2.840.113704.1.111.2744.1561686959.5 UI 38
[0020,0060] Laterality CS 0
[0020,1040] PositionReferenceIndicator LO 0
[0020,1041] SliceLocation 1781.5 DS 6
[0020,4000] ImageComments bone LT 4
[0028,0002] SamplesPerPixel 1 US 2
[0028,0004] PhotometricInterpretation MONOCHROME2 CS 12
[0028,0010] Rows 512 US 2
[0028,0011] Columns 512 US 2
[0028,0030] PixelSpacing [2] 0.48828125, 0.48828125 DS 22
[0028,0100] BitsAllocated 16 US 2
[0028,0101] BitsStored 12 US 2
[0028,0102] HighBit 11 US 2
[0028,0103] PixelRepresentation 0 US 2
[0028,1050] WindowCenter [2] 600, 600 DS 8
[0028,1051] WindowWidth [2] 1600, 1600 DS 10
[0028,1052] RescaleIntercept -1024 DS 6
[0028,1053] RescaleSlope 1 DS 2
[0032,1032] RequestingPhysician PET-11909-2 PN 12
[0032,1033] RequestingService LO 0
[0032,1060] RequestedProcedureDescription PET-11909-2 LO 12
[0032,1070] RequestedContrastAgent LO 0
[0038,0010] AdmissionID PET-11909-2 LO 12
[0038,0050] SpecialNeeds LO 0
[0038,0500] PatientState LO 0
[0040,0012] PreMedication LO 0
[0040,0253] PerformedProcedureStepID 5043144 SH 8
[0040,0275] RequestAttributesSequence SQ 120
[fffe,e000] Item na 104
[0040,0007] ScheduledProcedureStepDescription LO 0
[0040,0008] ScheduledProtocolCodeSequence SQ 40
[fffe,e000] Item na 24
[0008,0100] CodeValue SH 0
[0008,0102] CodingSchemeDesignator SH 0
[0008,0104] CodeMeaning LO 0
[0040,0009] ScheduledProcedureStepID PET-11909-2 SH 12
[0040,1001] RequestedProcedureID PET-11909-2 SH 12
[00e1,0010] PrivateCreator ELSCINT1 LO 8
[00e1,1001] DataDictionaryVersion 1 US 2
[00e1,1022] Unknown [2] 0, 0 DS 4
[00e1,1023] Unknown [2] 1, 1 DS 4
[00e1,103f] Unknown Tag & Data PETCT CS 6
[00e1,1040] OffsetFromCTMRImages bone SH 4
[00e1,1042] Unknown Tag & Data LO 0
[00e1,1061] Unknown Tag & Data LYPET_Multi_1002_usr.proc LO 26
[00e1,1063] Unknown Tag & Data CHINESE SH 8
[00e1,10c2] Unknown Tag & Data 1.2.840.113704.1.111.2728.1561687754.64 UI 40
[01f1,0010] PrivateCreator ELSCINT1 LO 8
[01f1,1001] Unknown Tag & Data SLICES CS 6
[01f1,1002] Unknown Tag & Data STANDARD CS 8
[01f1,1003] Unknown Tag & Data FUSED CS 6
[01f1,1004] Unknown Tag & Data HIGH CS 4
[01f1,1008] Unknown Tag & Data 168 DS 4
[01f1,100a] Unknown Tag & Data 0 US 2
[01f1,100c] Unknown Tag & Data [2] 0, 0.064000003 DS 14
[01f1,100d] Unknown Tag & Data 0 DS 2
[01f1,100e] Unknown Tag & Data 0 FL 4
[01f1,1027] Unknown Tag & Data 1.5 DS 4
[01f1,1028] Unknown Tag & Data 24 DS 2
[01f1,1030] Unknown Tag & Data 8 US 2
[01f1,1032] Unknown Tag & Data RIGHT_ON_LEFT CS 14
[01f1,1033] Unknown Tag & Data 3.3 DS 4
[01f1,1042] Unknown Tag & Data No SH 2
[01f1,1044] Unknown Tag & Data aaaa OW 644
[01f1,1046] Unknown Tag & Data 1.5 FL 4
[01f1,1047] Unknown Tag & Data 2D SH 2
[01f1,1049] Unknown Tag & Data 500 DS 4
[01f1,104a] Unknown Tag & Data NONE SH 4
[01f1,104b] Unknown Tag & Data 16x1.5 SH 6
[01f1,104c] Unknown Tag & Data NO SH 2
[01f1,104d] Unknown Tag & Data NO SH 2
[01f1,104e] Unknown Tag & Data Brain LO 6
[01f7,0010] PrivateCreator ELSCINT1 LO 8
[01f7,1010] Unknown Tag & Data 00 OB 2
[01f7,1011] Unknown Tag & Data 7c93 OW 488
[01f7,1013] Unknown Tag & Data 3a43 OW 136
[01f7,1014] Unknown Tag & Data 0000 OW 108
[01f7,1015] Unknown Tag & Data 02a0 OW 188
[01f7,1016] Unknown Tag & Data 772c OW 40
[01f7,1017] Unknown Tag & Data 0000 OW 8
[01f7,1018] Unknown Tag & Data 0000 OW 228
[01f7,1019] Unknown Tag & Data 0000 OW 2160
[01f7,101a] Unknown Tag & Data 0000 OW 28
[01f7,101b] Unknown Tag & Data 02a0 OW 1272
[01f7,101c] Unknown Tag & Data 0001 OW 116
[01f7,101e] Unknown Tag & Data 0000 OW 364
[01f7,101f] Unknown Tag & Data 0000 OW 148
[01f7,1022] Unknown Tag & Data 1.2.840.113704.1.111.2744.1561687837.19.1111.111111111111111 UI 60
[01f7,1023] Unknown Tag & Data 0016 OW 4
[01f7,1025] Unknown Tag & Data 0005 OW 12
[01f7,1026] Unknown Tag & Data f22d OW 13076
[01f7,1027] Unknown Tag & Data c000 OW 36
[01f7,1029] Unknown Tag & Data 4000 OW 440
[01f7,102b] Unknown Tag & Data 0008 OW 36
[01f7,102c] Unknown Tag & Data 0000 OW 656
[01f7,102d] Unknown Tag & Data 0000 OW 8
[01f7,102e] Unknown Tag & Data 0100 OW 128
[01f7,1030] Unknown Tag & Data 0000 OW 16
[01f7,1070] Unknown Tag & Data 0001 OW 584
[01f7,1074] Unknown Tag & Data 0000 OW 288
[01f7,1075] Unknown Tag & Data 0001 OW 136
[07a1,0010] PrivateCreator ELSCINT1 LO 8
[07a1,100a] Unknown Tag & Data 1e OB 282228
[07a1,1010] Unknown Tag & Data 3.5 LO 4
[07a1,1011] Unknown Tag & Data PMSCT_RLE1 CS 10
[7fdf,0010] PrivateCreator ELSCINT1 LO 8
[fffc,fffc] DataSetTrailingPadding 00 OB 392

2

Operating system: Mac Os
Slicer version:4.10.237%20PM
Expected behavior: load pet/ct data
Actual behavior: return as “Could not load: 2: Body-Low Dose CT as a Scalar Volume”
The pet/ct data is from Philips Gemini

3

The series are in private Elscint format with private transfer syntax CT-private-ELE, pixel data is in private data element, also compression in private PMSCT_RLE1. AFAIK, they shouldn’t leave Philips infrastructure, workstations can export files in standard DICOM for interchange, but from time to time they appear here and there. You can download converted files here. Please let me know if you have downloaded the files, they seems to be not anonymized and i’ll delete them after.
Best regards

1 Like
4

thanks, issakomi. that’s very helpful, I am new to image processing and not familiar with format. I will ask the tech about how to get standard DICOM. Is there any easy method to convert all my data to DICOM as you did? I have a bunch of data. Although the files have the patient name, they can not be located with only names here. However, it’s better to make the data full anonymized. Thanks for your reminder.

1 Like
5

one more question, why SUV is missing after I load the converted data to 3d slicer? thank you.

6

Is there any easy method to convert all my data to DICOM as you did?

There is rle2img.cxx example in GDCM’s Examples folder, i took it as starting point some time ago, but IMHO it seems to be broken after last commit at 19 May.

However, it’s better to make the data full anonymized.

One tip - if you will anonymize files - better don’t modify dates/time for PET series, if possible. If possible also preserving UIDs were good, but it is your decision, of course.
It is very easy to break PET series.

one more question, why SUV is missing after I load the converted data to 3d slicer? thank you.

The units in PET series are BQML (s. tag 0x0054, 0x1001), AFAIK SUV have to be calculated first. S. pdf files here. That conversion extracted pixel data from private data element and put it into standard data element (0x7fe0,0x0010) and set transfer syntax appropriately, all the rest incl. private tags is not modified, should be OK, but i can not guarantee, this is reverse-engineered stuff. BTW, i was able to create SUV from that PET series.

I have a bunch of data.

I could convert more series, if you like, it is anyway already done.
BTW, i am looking for example of Elscint files with PMSCT_RGB1 compression.

Best regards

7

@issakomi It would be nice if you could implement a DICOM reader plugin for this image type. You can use either Python or C++. If you are interested in this then let us know and we can give pointers to where to start.

8

This compression scheme is handled by the upcoming dcm2niix. Therefore, one can convert these with the SlicerDcm2nii extension. You can test this out already by using the developmental branch of dcm2niix.

2 Likes
9

However, it’s better to make the data full anonymized.

Again about de-identify and PET. I would recommend to use following options
from PS3.15 Table E.1-1. Application Level Confidentiality Profile Attributes

Rtn. Safe Priv. Opt. (PET related attributes are often private)
Rtn. Pat. Chars. Opt. (Patient Weight is important)
Rtn. Long. Full Dates Opt. or Rtn. Long. Modif. Dates Opt (better 1st, retain full, even smallest error will break things)
Rtn. UIDs Opt. (Frame of Reference UID may be important for fusion later)

or just hide name, birth date, institution, etc., if possible

10

@lassoan this is a very simple but not very effective compression format. The GDCM C++ code is here, or if you prefer the dcm2niix C function nii_loadImgPMSCT_RLE1() is here. Both are based on the BSD License.

2 Likes
11

@issakomi thanks for your reply, the information very helpful.
There are around 350 datasets to be converted, so it’s better for me to do it offline, I am very fresh to coding, AKA. layman, I don’t know how to employ pure coding yet. Could the convert to be done by click some kind of button or special software? all my data are in the same format as I uploaded before.
For SUV part, I will study the files you direct. Honestly, it is a lot for me to digest, but I’d like to try. So glad to learn new things.
PS. I am a radiologist lack of engineering background.

1 Like