The DICOM Identifiers and the Concept of the DICOM Study The DICOM Identifiers and the Concept of the DICOM Study

The most important concepts of the DICOM standard are:

 the study and the unique study identifier

 and other identifiers associated with series

 and images and

 identifiers related to another DICOM object that will not be discussed here.

Think of an old fashioned X-ray imaging center. The film based images belonging to an examination of a given patient would be put into an envelope, also all of the images would bear some of the patient data and further information like the image orientation. What makes up a study, is that the images belonging to an examination are kept in the same envelope, and on the envelope we have information sufficient to identify the examination, i.e. the images and the patient. If by accident some images fall out from envelopes then based on the patient data appearing on all of the images and on the envelopes, we could restore the original order.

That is, a study is a set of well identifiable images belonging to a well defined envelope: All the images contain information about the envelope, and the envelopes contain information about the images. From the point of view of the order of the imaging center, the simple fact that a poor patient is also involved is unimportant compared to the need that images and envelops can be uniquely associated with each other.

74

Image

Now imagine a state of the art digital imaging center operating with modalities creating hundreds of digital images at an examination. Where are the envelopes and where is the information that will help us to put together the images belonging to the same examination, the study? First of all we learned from the old fashioned imaging center that all the images will have to contain the same information in order for us to be able to restore the study if fallen apart. In the case of studies containing only a few images it seems to be easy to restore the order. The same task may be may be very hard to accomplish in the case of studies of hundreds of images. However, computers are very good at numbers. So, take a unique numeric ID that will identify the study and attach that number to all of the images belonging to this study. Then the computer will know which images belong to which study. This is the state of the art version of the old envelope-images system. So, in DICOM the envelopes get immaterialized and a system of unique study identifiers replaces them.

If we have a unique study UID, a Unique ID associated with one and only one study on earth then inside computer we can keep order. We repeat that the study ID has to be unique. It is only the DICOM organization that can issue identifiers for companies manufacturing imaging equipment and developing software that may modify DICOM studies. This is how a study UID looks like:

1. 3. 12. 2. 1107. 5. 1. 4. 45533. 30000008102706401187500000052

If we compared this study UID with other study UID‟s generated by the same device we would see that the first digits of the identifier would be the same and from a certain digit the numbers are different. This is because the first half of the study UID belong to the manufacturer and the specific device and the last characters identify the examination. Further to the Study UID that appear in every images, all the series and all the images have their own unique ID.

The DICOM standard may seem to overkill the identification procedure because not only the studies but the images and the series have their own ID‟s. In the case of the series having their own ID contained in every image belonging to them is not an overkill: the machine can always reproduce the series like they were contained in sub envelops of the great envelop containing the study. But why do images have their own ID? Think of a modality resending some of the images of a study. If the ID is the same then the new image simply replaces its previous version. If the „resent” images have new ID‟s then they will be handled as new images to the study whose ID is also contained in the resent images.

This system of unique ID‟s assure that studies, series and images will never be mistakenly replaced or mixed.

75 2.5.4.3 The structure of a DICOM file, the DICOM tags

As we already know a DICOM file storing one image contain the image data and data belonging to the patient and data (name, age, etc.) belonging to the examination (date of acquisition, manufacturer, etc.) and identifiers: the study UID, the series‟ UID‟s, and the image UID‟s.

The software that interprets the image will have to be able to find, first of all, the part of the DICOM file containing the image; also all of the identifiers and the other data contained in the DICOM file. The DICOM standard has a special pair of characters, the parentheses and the comma: ‟(‟ and ‟)‟ and ‟,‟. Now, numbers of 2x4 hexadecimal digits enclosed by the these parentheses and separated by the comma uniquely identify a specific DICOM field or data.

For instance this tag:

(0010,0010)

is the identifier of the patient‟s name - „ten-ten is the patient name” as DICOM experts would say. The last thing that we have to learn is that the data, in this case the patient name is enclosed by a pair of the tag shown above:

… (0010,0010)TEST^EDWARD

Here ^ is another special character separating certain sub data between the same tags. When the computer finds the first ( character then it is ready to learn the tag identifier given before the next ) character. When he machine finds the closing tag within the closing pair of parentheses it will check if the same tag is found. If not then there must be an error in the DICOM file. If no error is found then one DICOM data like the patient name is learned. After decoding all the data like the patient‟s birth date and the part containing the actual image, the computer is able to interpret the result as an image on the screenl.

Here is a decoded segment of the DICOM information found in a DICOM file:

1. Dicom-File-Format

1. Dicom-Meta-Information-Header

2. Used TransferSyntax: LittleEndianExplicit (0002,0000) UL 182 # 4, 1 MetaElementGroupLength (0002,0001) OB 00\01 # 2, 1 FileMetaInformationVersion

(0002,0002) UI =CTImageStorage # 26, 1 MediaStorageSOPClassUID (0002,0003) UI 1.3.12.2.1107.5.1.1.20377.20031125114113176.4 # 46, 1 MediaStorageSOPInstanceUID

(0002,0010) UI =LittleEndianImplicit # 18, 1 TransferSyntaxUID

(0002,0012) UI 1.2.826.0.1.3680043.2.93.0.99 # 30, 1 ImplementationClassUID (0002,0013) SH ERAD_60 # 8, 1 ImplementationVersionName

76

1. Dicom-Data-Set

2. Used TransferSyntax: LittleEndianImplicit

(0008,0005) CS ISO_IR 100 # 10, 1 SpecificCharacterSet

(0008,0008) CS ORIGINAL\PRIMARY\LOCALIZER\CT_SOM4 TOP # 38, 4 ImageType (0008,0016) UI =CTImageStorage # 26, 1 SOPClassUID

(0008,0018) UI 1.3.12.2.1107.5.1.1.20377.20031125114113176.4 # 46, 1 SOPInstanceUID (0008,0020) DA 20031125 # 8, 1 StudyDate

(0008,0021) DA 20031125 # 8, 1 SeriesDate (0008,0022) DA 20031125 # 8, 1 AcquisitionDate (0008,0023) DA 20031125 # 8, 1 ContentDate (0008,0030) TM 113945.000000 # 14, 1 StudyTime (0008,0031) TM 114003.384000 # 14, 1 SeriesTime (0008,0032) TM 114109.299000 # 14, 1 AcquisitionTime (0008,0033) TM 114109.299000 # 14, 1 ContentTime (0008,0040) US 0 # 2, 1 ACR_NEMA_OldDataSetType

(0008,0041) LO IMA TOPO # 8, 1 ACR_NEMA_DataSetSubtype (0008,0050) SH (no value available) # 0, 0 AccessionNumber (0008,0060) CS CT # 2, 1 Modality

(0008,0070) LO SIEMENS # 8, 1 Manufacturer

(0008,0080) LO SE-AOK RAD.ONKOT.KLIN # 22, 1 InstitutionName (0008,0090) PN FORGACS # 8, 1 ReferringPhysiciansName

(0008,1010) SH sict04 # 6, 1 StationName

(0008,1030) LO NATIV DR. FORGACS/ TA # 42, 1 StudyDescription (0008,1090) LO SOMATOM PLUS 4 # 14, 1 ManufacturersModelName (0009,0010) LO SPI RELEASE 1 # 14, 1 PrivateCreator

(0009,0012) LO SIEMENS CM VA0 CMS # 20, 1 PrivateCreator (0009,0013) LO SIEMENS CM VA0 LAB # 20, 1 PrivateCreator (0009,0020) LO SIEMENS CT VA0 IDE # 20, 1 PrivateCreator (0009,0030) LO SIEMENS CT VA0 ORI # 20, 1 PrivateCreator (0009,1010) LT SPI VERSION 01.00 # 18, 1 Comments

The first column contains the DICOM tags, the second column contains the actual content, the third column contains the actual number of characters belonging, the fourth column tells if the actual field has more than one data contained (these are back slash separated like in the second row) and the official DICOM name of the field.

The DICOM standard has specific names for the different DICOM tags or data or DICOM field identified by a certain pair of tags. For instance (0008,0060) is the tag identifying the DICOM field called Modality. The content of this field above is CT. Another example is the StudyDate being the official name of the field containing the actual date of the study: the tag is (0008,0020) and the content above is 20031006, that is November 6, 2003. As an exercise, find the tag and the DICOM data describing the manufacturer f he actual modality. What is the tag? What is the official DICOM name? How many characters do we have? Who is the manufacturer?

(1) The most important DICOM information is the sequence of numbers describing the image itself. The image information is also introduced by an identical tag. Since there are different methods for the numeric interpretation of the pixels of an image, the DICOM information

77 must contain some data about the said code. We are not going to sink into further details regarding the different numeric image representation methods.

However, there are other DICOM data that the computer needs to find when actually rendering an image. In the following paragraph we are going to learn about the methodology DICOM is using when describing the spatial location, orientation and the physical size of an image.