1 ***************************
3 [Correlating CT and PT series]
4 http://groups.google.com/group/comp.protocols.dicom/browse_thread/thread/50bd859e0232a735/6040b545240c8fa4
6 For completeness, there are three ways an application can
7 correlate CT and PET (or other sets) using DICOM only:
9 1) Code all the images with the same Frame of Reference UID.
10 At least one major vendor does this now.
12 2) Make assumptions based on specific implementation behavior
13 based on data elements that match (e.g. perhaps Study Time, or
14 matching values in the General Equipment Module).
15 Without actual trial and experience verifying the correlation,
16 I agree, that an assumption is not wise
18 3) Use the recent Supplement 73 Spatial Registration Storage SOP Classes.
19 This capability allows explicitly specifying the spatial transformation
20 from one Frame of Reference to another. However, this supplement
21 is probably too recent to be in use.
23 [Is there a SOP class for CT-PET images?]
24 http://groups.google.com/group/comp.protocols.dicom/browse_thread/thread/bb96abb730624a8a/5a78f1ff0199a727
26 [Everything you have always hated about the DICOM PET SOP Class, but were afraid to ask]
27 http://groups.google.com/group/comp.protocols.dicom/browse_thread/thread/dc3a07d4999445ae/560fd0a064145d61
30 http://groups.google.com/group/comp.protocols.dicom/browse_thread/thread/30897c138442daad/27382b21968ce0b3
32 WG 17 is considering the matter of registration of images
33 in 3D, including rigid (and probably non-rigid) body
34 transformations. This effort has been prioritized over
35 WG 17's other work on multi-dimensional representation
36 as a consequence of the prevalence of PET-CT fusion
39 ftp://medical.nema.org/medical/dicom/supps/sup73_05.pdf
40 broken link use instead: ftp://medical.nema.org/medical/dicom/final/sup73_ft4.pdf
43 http://groups.google.com/group/comp.protocols.dicom/browse_thread/thread/56d07d5686e2ec67/b4c131f3e711716f
46 ***************************
47 - Spatial registration oriented:
48 [Spatial Registration]
49 http://groups.google.com/group/comp.protocols.dicom/browse_thread/thread/ab94a61d33082e29/d929474e92c30b24
51 [presentation context ID]
52 http://groups.google.com/group/comp.protocols.dicom/browse_thread/thread/a4bb2647d5a91001/d0fbff58a644ce7c
54 ***************************
55 Adding comment by Doug Sluis (dsluis at clinical-knowledge.com)
56 To your numbered questions:
57 1. Yes they are separate SOP Instances. (What is alternative?)
58 2. Yes. (although the recent Spatial Registration SOP Class is a possible option)
60 There is no guarantee that CT and PET slices are one-to-one or
61 that the slices are spatially coincident. Attempts to spatially sort
64 > I am trying to gather some information about CT-PET acquisition and how
65 > to order/link slices together properly(*).
67 > 1. I understand there is not and there won't be any SOP Class for
68 > CT-PET. But should we still consider this as two *independent* SOP
71 > 2. Image acquired in this dual modality *have* to have the same Frame
72 > of Reference UID, right ?
74 > In this case I'll add the following strategy in gdcm to organize the
77 > 1. Open a DICOM file A
78 > 2. Open another DICOM file B
79 > 3. Same Serie/Study ?
82 > 4. Same Frame of Reference ?
85 > 5 A is CT and B is PT ?
88 > 6 Same Image Position ?
90 > * Yes: We found a match !