Introduction to DICOM. Characteristics
The DICOM format
In medical imaging there is a worldwide recognized standard format, which is [DICOM] (Digital Imaging and Communication in Medicine).
This format has been formulated for the handling, visualization, storage, printing and transmission of medical images. And the organization primarily responsible for maintaining this standard is the [NEMA] (National Electrical Manufacturers Association). Initially the protocol was defined by the American College of Radiology (ACR) and the NEMA in 1985 and meets the following requirements:
- It is applicable to network and offline environments (independent computers, for example).
- Specifies how all protocol-compliant devices should react to commands and data exchange.
- Specifies the different levels of compliance.
- Gives the structure of a multi-part document.
- Gives explicit information not only for images but also for reports, print formats, etc.
- Specifies a way for the unique identification of any piece of information (object).
It includes the definition of a file format and a network communication protocol (applies the TCP/IP protocol). And it allows different entities to share information and keep it usable and legible.
DICOM allows the integration of scanners, servers, workstations, printers and network hardware from multiple vendors into an image storage and communication system. The different machines, servers and workstations have a DICOM declaration of conformity (conformance statements) that clearly establishes the DICOM classes they support. DICOM has been widely adopted by hospitals, managed through systemsPACS.
Structure of a DICOM file
The structure of the DICOM file is the following (Figure 1 it is taken from the web of [leadtools]:
Header
This part of the file can be omitted, although the standard specifies that the file must include at least one header. This section is composed of:
- Preamble: 128 bytes and is usually used to indicate multimedia applications access to one of the images within the DataSet that will serve as an icon or visual identification of the same.
- Prefix: of 4 bytes, contains the string DCIM, in capital letters.
DataSet
Each file will contain a single DataSet, consisting of several DataElements.
The DataSet consists of a SOP pair (Service-Object Pair), these pairs relate an IOD (Information Object Definition) and a DICOM Service Elements (DIMSE). An example of SOP is: Store, Get, Find, Move. A DIMSE can be a CT image, MRI or even agendas (schedules) and other data in text format; DIMSE can be type N (normalized, sólo para servicios de notificación) and type C (for operation and notification services). For more information on PCOS, see [here]. Based on the two types of DIMSE, there will be two SOP classes: normalized (normalized IOD + DIMSE-N) or composite (composite IOD + DIMSE-C).
DataElements store the values of the coded attributes for the real-world object represented in the DataSet. Each DataElement is identified by its tag (DataElement Tag), these tags are numbered with a unique identifier and ordered within the DICOM file from minor to major.
Contents of a DICOM file
Within each DataSet, information about the patient (name, age, administrative data, etc.), date on which the data were taken, data format, model of the device used to record the data (some of the fields added are manufacturer-dependent, but always in accordance with the DICOM format), etc., will be stored within each DataSet. For the same patient there will be different DataSets, depending on the number of tests that have been carried out over time. This allows you to view your file, for example, within a PACS server, both locally and networked.
