italiano
english
 INTRODUCTION TO DICOM STANDARD
version 1.0
  

Brief introduction
The DICOM format (Digital Imaging and COmmunications in Medicines) is a standard (i.e. a set of rules) for transmitting digital data between different devices (scanner, computer, printers, digital archievies, etc.) like biomedical images and health information through the establishment of a communication protocol and the use of technology and standard network protocols (Ethernet, TCP/IP). This standard is a protocol to application level (according to the ISO/OSI model), it uses the TCP/IP net protocol to make to communicate between the systems. The DICOM files can be exchanged among two entities that possess the capability to receive the information (images and data of the patient) in the DICOM format.
The DICOM format is public, in the sense that the definition of the standard is accessible to everyone. It reveals extremely advantageous, allowing to store, to elaborate and to exchange digital images in efficient way among different devices produced by different company.
 
Brief history
The native project was developed by two American associations: The American College of Radiology (ACR), responsible of the technical-medical development of the system and the National Electrical Manufacturers Association (NEMA), responsible consortium of producers besides of the inherent aspects to possible violations of brevets and normative. To international level in Europe, the committee of standardization has acknowledged the DICOM format in MEDICOM; as the JIRA (Japan Industries Association of Radiological Systems) has approved the development of this. In 1985 come officialized the version 1.0 of the standard ACR-NEMA to which followed in 1988 the version 2.0: it dealt with a primitive standard in which the format of the that containing the files was defined the images and the physical standard and of protocol for the interconnection pixel-stung of the various equipments. The implementations nevertheless they were rather limited, above all because of the mean physicist of connection already realized with technologies for the epoch obsolete. In 1993 the standard ACR-NEMA transformed him radically in the version 3.0 in which, substantially maintaining unchanged the inherent specifications the format of the images, was assistant manifold services and implemented the protocols of net TCP/IP and OSI: The new standard was identified with the term DICOM, and really the integration in the specifications of the protocol of net TCP/IP, by now largely diffused, it always decreed a success and a popularity of it increasing.
 
Net DICOM services
The net services of DICOM are based on the client/server concept. In the case two DICOM applications want to exchange some information, it is necessary to establish a connection and an accord on the following parameters:
  1. who is the client and who is the server;
  2. what DICOM services must be used;
  3. in what format the data are transmitted (i.e. compressed or not compressed).
Only if both the applications agree on this whole parameters the connection can happen. Additionally to the most fundamental service DICOM, related to the transmission of images (or in terminology DICOM: "Storage Service Class"), a certain number of advanced services are available, for instance:
  • The service of images storage ("Query / Retrieve Service Class") allows to seek images in a PACS archieve according to certain criterions (patient, image acquisition date/time, equipment, etc.) and to selectively download the images from this archieve.
  • The print DICOM service ("Print Management Service Class") allows to access to the network printers, so that different equipments and workstation can share a printer.
  • The modality worklist service DICOM allows to automatically download some refreshed worklists, comprehensive of demografic data of the patients, from the informative systems type HIS/RIS toward the equipment (modality) client.
 
DICOM files format
A file DICOM in its essence can be seen how a container, this standard doesn't introduce any new format for the images (as the types jpeg, gif etc.), the contained data don't only represent the pixels and the colors of the image, but also a set of structured data that describe the whole procedure that has brought to the construction of the same image. Substantially a file DICOM consists therefore of a header constituted by an attributes set containing information of various nature and by a data body to contain one or more images. The data image can be also compressed through an ample variety of standard compression algorithms of images, among which JPEG, LZW and Run-Length Encoding (RLE). The attributes set of the header can be grouped according to the relationship that exists between them, and they goes to form that entity known as Information Object Modules, then every group can be identified as the abstraction of a real entity, that can be for instance: Patient, Study, Series and Image (in figure 1 are possible to see a first simplification as a file DICOM is structured). In the first group of attributes there are present information such as the personal data of the patient (name, Id, birthday, sex, etc.) submitted to medical exam; in the second group there are present the the characteristics of different analysis methods (modalities) which constitute the diagnostic study (date, time, technician, etc.); in the third group, defined series, are collected the data describing the collections of images from each modality with the relative parameters of acquisition (number of the series, modality type, etc.); finally the quarter and last group is that contains the descriptive attributes of the images as the array dimension, the depth of the pixel, the photometric interpretation etc. Naturally these are not the only groups present, but surely those fundamental for a simple introduction.
As we said the standard DICOM mainly differs from other data formats because it allows to encode the information in a set of attributes (data set) that represent an informative instance. Then the attributes can be seen as the fundamental building blocks of a DICOM file, structurally an attribute is formed by a fields set:
  • the unique name;
  • an unique tag (label);
  • a semantic description;
  • a value representation;
  • a value multiplicity.
figure 1
 
From this list has been omitted, intentionally, for simplicity the field type classification. The field name has a pure mnemonic functionality and serves to the human being to identify more easily the attribute. The field tag always serves for identifying the attribute, but being constituted by two integers it results easier to treat from the informative systems; these two numbers are called group number and element number; they are canonically represented by two hexadecimal numbers of four digits separated by comma and contained among round bracket, i.e. the couple (0010,0020) identifies the ID Patient. The description, as it is easy to realize, it furnishes through a sentence the meaning of the attribute. The field value representation (constituted by a string of two characters) specifies the data type and the format of the value represented in the attribute. The field value multiplicity points out the number of values contained in the attribute, can also point out a varying quantity, for instance "1-10" points out that in the attribute there can be from 1 to 10 values.
 
For example an image of a thoracic radiography will be present in the same file with the patient id, so that the image is never separated by the other personal data. The presence of the file is also recommended DICOMDIR that furnishes the indexes and a table of contents of the data of the all DICOM files present on the media storage.