DICOM standards and applications-Introduction to DICOM standards

1. What is DICOM?
DICOM, short for digital imaging and communication of medicine, is an American College of radiology (ACR) and National Electrical Manufacturers Association (SCN) the Organization's standard name specifically for the storage and transmission of medical images. After more than a decade of development, this standard has been widely accepted by medical equipment manufacturers and the medical industry and has been widely used in medical instruments. computer tomography (CT) with DICOM Interfaces) magnetic Resonance (MR), cardiovascular angiogram and ultrasound imaging equipment appear in large numbers, and play an important role in digital network of medical information systems.
DICOM is a digital and computer-based medical equipment and hospital management information system, especially the image archiving and communication system (PACS) and the development of remote medical systems came into being. When CT, Mr, and other devices generate high-quality, visually visualized images that are widely used in medical diagnostics, images produced by devices of different manufacturers and models are in different formats, it makes information resources between different devices difficult to use each other, and the implementation of hospital PACS is very difficult. Medical Information systems bring about many new problems: how to store images with a large amount of data and effectively manage them? Can devices of different manufacturers be connected directly? How can information resources be shared between devices of different manufacturers? And so on. Obviously, the solution to these problems is to adopt unified standards. To this end, the American radiology Society and the American Electrical Manufacturers Association established a special committee in 1983 to develop standards for medical image storage and communication, provides unified formats of digital images irrelevant to manufacturers and their associated communication and storage functions to Promote the Development of PACS and provide a wide range of distributed diagnostic and query functions. The ACR-NEMA1.0 version was launched in 1985, followed by new data elements and modifications to some of the content to form version 2.0. Due to the recognition of the lack of network support standards and the structural problems of the standards themselves, the ACR-NEMA combined with the current technical conditions and methods to completely reformulate the standards, A new version named dicom3.0 was officially released in 1993. Compared with the original version, version 3.0 uses an object-oriented analysis method to define various entities and relationships of medical images during storage and communication, supports ISO-OSI (Inter-national standard organization-Open System Interconnection) and TCP/IP (Transmission Control Protocol/Internet Protocol, in this way, the medical image application layer can communicate directly with other communication protocol stacks without re-writing the program. Considering the development of technology, the standard adopts a multi-part document structure, which is provided in the form of an Appendix for the parts that may be changed or expanded, so that the standard involves as little as possible during the update.
Basic concepts and definitions involved in standard 2
The DICOM standard involves medical image, data communication, management information system and other fields. In the standard, the object-oriented description method and E-R (entity-relation) model are adopted, as a result, a large number of professional terms are introduced, which brings difficulties to the standard reading and understanding. Below we will briefly explain the frequently used technical terms and acronyms involved in the standard.
1. entity: indicates one or more application objects with the same characteristics. In computer system analysis, all things, things, and concepts that can be differentiated and recognized by people can be abstracted as entities. An object has several characteristics, which are called attributes. For example, a patient is an entity with attributes such as name, gender, and age. An image is also an entity that has attributes such as image size and image data.
2. Relation: the relationship between entities. If there is a reference link between the patient entity and the analysis entity, there is a print link between the printer entity and the film entity.
3. E-R model: an information model describing the real world. Express system requirements and functions by defining the relationships between entities and entities. It is typically represented as a E-R graph. In DICOM, a square represents an object, a diamond represents a link, and an arrow or a line segment without an arrow connects an object (box) to a contact (diamond) to indicate a link between them. This is the main representation method used by object-oriented analysis methods and an abstraction of the objective world.
4. Object: the representation of the external world in the computer. It is a collection of the attribute values and processing methods of the transaction. Objects have encapsulation and inheritance features. Encapsulation refers to the combination of attributes and methods of an object, which is generally provided only to you and the derived object. Inheritance means that when an object is derived from another object (parent object), it automatically has the attributes and methods of the parent object. The object-oriented method is centered on object technology. It analyzes the relationship between various information in the system, abstracts object models at various levels of the system, and gives accurate system descriptions, and implement it in the computer system. The application object-oriented method can improve development efficiency and achieve software reuse.
5. Information ob-ject definition (IOD): Abstract of information objects, which is the role receptor of DICOM commands.
6. Service: functions provided by an object for other objects or programs. When this function is required, it is called the application service. The application object is called the service user, and the object that can complete this function is the service provider.
7. service object pair (SOP): Basic Function Unit for transmitting DICOM information. Contains an information object and a group of DICOM message service elements.
8. Protocol: communication rules and formats that must be followed in computer networks to ensure correct data transmission.
9. ISO-OSI: A layer-7 network reference model for Open Systems Interconnection (OSI) defined by the International Organization for Standardization (ISO. As a strict network model, it plays an important role in the research and development of computer networks. However, due to various reasons, it has not been widely used in practice. During the preparation of DICOM standards, OSI is the climax of development. Therefore, it is also the main network Reference Model in DICOM.
10. TCP/IP: the transmission control protocol and Internet protocol. It is first used in UNIX systems and then becomes the main communication protocol for communications between different types of computers in computer networks. It is the basis of the Internet.
Composition of the three standards
DICOM standard is a process of development from scratch, from simplicity to complexity. Constantly listen to comments and suggestions from the industrial, academic, and medical sectors during the formulation of standards, and pay attention to the scalability and scalability of standards, having gone through versions of ACR-NEMA 1.0 and 2.0 to the current DICOM 3.0, the standard composition is constantly being supplemented. Currently, the standard consists of the following 14 Basic Parts and expansion parts, as shown in Figure 1:
1. Part 1: Provides the standard design principles, defines some terms used in the standard, and gives a brief overview of other parts of the standard.
2. Part 1: definitions and methods of DICOM compatibility are provided. Compatibility refers to the ability of devices that comply with DICOM standards to connect and operate on each other. Due to the large content and complex functions of the DICOM standard, there are currently no devices that can cover all DICOM functions, but are only necessary to implement the functions of this device. Therefore, the device manufacturer must provide a description of the DICOM functions supported by the device, that is, the Compatibility Statement. The standard content of this section defines the declared structure and information that must be presented, including three main parts:
A. A set of information objects that can be identified in this implementation;
B. A set of service classes supported by this implementation;
C. A set of communication protocols supported by this implementation.
The standard does not specify the testing and verification process for compatibility implementation. When purchasing DICOM devices, you must check whether the compatibility level of each device is consistent. Otherwise, some problems may occur during interconnection.
3. Part 1: describes how to define information objects and Provides abstract definitions for Information Objects in medical digital image storage and communication. Each information object definition consists of its purpose and attributes. To facilitate standard expansion and maintain compatibility with old versions, information object classes of compound and common types are defined in DICOM. Common Information object classes only contain the attributes inherent in real-world entities. Compound information object classes can be appended with properties that are not inherent in real-world entities. For example, the CT image information object class includes attributes of image entities, such as the Image Date and image data, and does not belong to the image itself, such as the patient name. The composite object class provides a structural framework to express the needs of image communication, making applications in the network environment more convenient.
4. Part 1: Description of the service class. The service class associates the information object with the command acting on the object, and describes the requirements of the command element and the results of the command acting on the information object. Typical DICOM services include query/search services, storage services, and print management services. A service class can be simply understood as a command provided by DICOM or an internal call function provided to an application. This part actually shows the command stream in DICOM messages.
5. Part 1: Data Structure and semantics: describes how DICOM application entities construct dataset information derived from the use of information objects and service classes, and provides data stream encoding rules for message transmission. Data streams are generated by data elements of a dataset. Several datasets can be referenced or tolerated by a composite dataset. A composite dataset can transmit information object content in a "data packet. This part focuses on the data flow in DICOM messages. In addition, it defines the semantics of common basic functions of many information objects, such as required conditions, completed results, and implemented functions.
6. Part 1: Data dictionary, which is a set of definitions of all data elements that represent information in DICOM. Unique tags, names, Digital features, and semantics are specified for each data element in the DICOM standard. In this way, when messages are exchanged between DICOM devices, the content in the message has clear and unambiguous numbers and meanings, which can be understood and interpreted by each other.
7th Part 1: message exchange. A message is composed of one or more commands for exchange and the data necessary to complete the command. It is the basic unit for communication between DICOM application entities. This section describes the services and protocols used by application entities for message exchange in the medical image environment.
8. Part 1: Network Support for message exchange. This section describes the communication service and necessary upper-layer protocol support between DICOM entities in the network environment. These services and protocols ensure the effective and correct communication between application entities over the network. The network environment in DICOM includes two reference models: OSI and TCP/IP. DICOM only uses rather than implements these two types of protocols, so it is universal.
9. Part 1: point-to-point communication support for message exchange. Describes services and protocols in point-to-point communication environments compatible with ACR-NEMA2.0. It includes the physical interface, the signal contact process, and session/transmission/network protocols and services similar to OSI that use the physical interface.
10. Part 1: media storage and file format for media exchange. This section describes a general model for storing medical image information on removable storage media. It provides a framework for exchanging different types of medical images and related information on various physical storage media, and supports encapsulating file formats defined by any information object.
11. Part 1: media storage application file, Compatibility Statement for information exchange between medical images and related equipment. The application description of images such as cardiovascular angiogram, ultrasound, CT and MRI and the exchange of CD-R format files are given.
12. Part 1: physical media and media format used for media exchange. It provides information exchange between Digital Image Computer Systems in medical environments. This switching feature will enhance diagnostic images and other potential clinical applications. This section describes the structure of the relationship between the media storage models and the specific physical media characteristics and their corresponding media formats. Detailed description of various specifications of the magnetic disc, PC using the file system and 1.44m floppy disk, as well as CD-R can be engraved CD.
13. Part 1: Print management supported by point-to-point communication. Defines the necessary services and protocols that support DICOM printing to manage application entity communication when printing point-to-point connections between users and print providers. The point-to-point communication file provides the same Upper-layer services as Part 1. Therefore, the print management application entity can be applied to point-to-point connections and network connections. Point-to-Point print management communication also uses a low-layer protocol, which is compatible with the existing parallel image channel and serial control channel hardware hard copy communication.
14. Part 1: Describes the standard display function of grayscale images. This section only provides methods for measuring the display characteristics of a specific display system. These methods can be used to change the display system to match the standard grayscale display function or to measure the compatibility between the Display System and the standard grayscale display function.
Application 4
There is no doubt that DICOM is the core of medical image information systems. It mainly involves the storage and communication of the most important and difficult medical images in information systems, it can be directly applied to Radiology Information Systems (RIS) and image archiving and communication systems (PACS. DICOM is also the technical basis for researching and developing new medical instruments with the network connection function and sharing information resources. While developing towards automation and intelligence, medical instruments are also developing towards remote control and telemetry with communication capabilities and network functions for Remote Information Retrieval. Medical instruments are not only the source of information in the medical information system, it is also an information user in the system and a major link in the Information System. The importance of networked medical instruments to medical information systems is self-evident.
Another feature of the DICOM standard is that it is defined at the top layer of the network communication protocol and directly applies the network protocol without specific hardware implementation. Therefore, it remains relatively independent from the development of network technology, the DICOM system performance can be improved immediately with the improvement of network performance. Although DICOM provides the OSI network model, most of the networks are constructed under the TCP/IP protocol. The network hardware can be in a variety of formats, such as m twisted pair wire 100base-t, optical fiber FDDI, Integrated Service Digital Network ISDN, T1 line, and other 10 Mbps network 10base-t and telephone lines with low speed. As long as the device has a network interface that supports the TCP/IP protocol, with the support of software, you can achieve "Plug and Play" like a PC ", it is very convenient to join the medical information system network. In this sense, the Medical Information System Implemented by DICOM, whether it is RIS or PACs, has a similar structure, as shown in Figure 2:
In an information network system that adopts DICOM standards, all DICOM devices can connect to each other and operate according to the DICOM upper-layer network protocol. Clinicians can view the images and results of B-ultrasound devices in the office, call MRI images on the CTF for overlapping and fusion, or call images stored in other hospitals over the Internet. DICOM devices can communicate with each other over the network to exchange information.
By providing a Unified Storage Format and communication mode and popularizing DICOM standards, the design of medical information systems can be simplified, avoiding a lot of repetitive work and accelerating the development of information systems. The implementation of paperless and non-film Hospital and remote medical systems will play an extremely important role.