This chapter includes the following sections:
Digital Imaging and Communications in Medicine (DICOM) is a medical imaging standard. This standard was initiated by the American College of Radiology (ACR) and the National Electrical Manufacturers Association (NEMA) to enhance the connectivity of radiological devices. Before the DICOM standard became widely adopted, each manufacturer had its own proprietary image format and communication protocol. This proliferation of formats and protocols made it almost impossible to produce third-party software to manage or study medical content. Nor was it possible to connect hardware devices from different manufacturers.
In 1985, the American College of Radiology and the National Electrical Manufacturers Association jointly published a medical imaging and communication standard, named the ACR-NEMA standard, to address this problem. In 1993, the ACR-NEMA standard was revised and renamed as DICOM (Version 3.0). Since then, the DICOM standard has become the dominant standard for radiology imaging and communication. All major manufacturers conform to this standard. Today, any software component can take DICOM content from any manufacturer and manage that content with a uniform interface. The term DICOM content refers to standalone DICOM Information Objects that are encoded according to the data structure and encoding definitions of PS 3.10-2007 of the DICOM standard (commonly referred to as DICOM Part 10 files). DICOM content can include different types of data, such as patient administration information, waveforms, images, slices of 3-D volumes, video segments, diagnostic reports, graphics, or text annotations. DICOM content contains a number of standard attributes. Optionally, DICOM content can also contain private attributes. (In the DICOM standard, the phrase DICOM objects refers to DICOM content.)
Like other standards, DICOM is mostly developed by volunteers. Working groups formed by domain experts propose additions and changes to the existing standards, and the changes are approved by a balloting process. Typically, NEMA publishes a new version of the standard each year. This standard is available worldwide from the NEMA Web site at
The DICOM standard has two major areas of focuses: the data model (or file format) and the communication protocol. The data model is defined using object-oriented programming principles. Content such as images and waveforms captured by medical devices are represented as information objects. Services such as get, find, and store operations can be defined on these objects. The services and the information objects are combined into a service object pair (SOP).
DICOM defines different types of transfer syntax (or binary encoding rules) for sending objects across the network or encoding objects in files. Transfer syntax specifies the mapping of a DICOM object hierarchy into a binary stream. The binary data can be stored on physical media such as tapes, CDs, or disks, and organized in accordance with the DICOM file hierarchy. The binary data can also be exchanged over a network with the DICOM communication protocol, which covers the upper layers (application, presentation, and session layer) of the Open Systems Interconnection (OSI) seven-layer model. The DICOM communication protocol is typically implemented on top of TCP/IP. Recently, the DICOM standard introduced Web access to DICOM objects (WADO). WADO deals primarily with HTTP access to DICOM objects. Messages exchanged between a DICOM server and a DICOM object involve operations such as radiology workflow, gray-scale image rendering, image printing, and storage and retrieval.
The Oracle Multimedia DICOM feature is concerned with the storage, management, and manipulation of DICOM format medical images and other objects encoded into files. Oracle Multimedia does not support the DICOM communication protocol.
The Digital Imaging and Communications in Medicine (DICOM) feature was first introduced to Oracle Multimedia in Oracle Database 10g Release 2 (10.2). For that release, Oracle Multimedia DICOM enhanced the previous behavior of the Oracle Multimedia ORDImage object type by allowing Oracle Multimedia to recognize DICOM content and extract a subset of embedded DICOM attributes relating to patient, study, and series.
Oracle Database 11g Release 1 (11.1) continues to provide the same DICOM support in ORDImage. In addition, this release provides more complete DICOM support in a new ORDDicom object type.
With Oracle Database 11g Release 1 (11.1), Oracle Multimedia provides full support for DICOM, the format universally recognized as the standard for medical imaging. Applications can now use Oracle Multimedia DICOM Java and PL/SQL APIs to store, manage, and manipulate DICOM content.
Customers can build large archives of medical content that are managed and secured using Oracle Database. Complete DICOM metadata support enables customers to index and search the archived DICOM content for research purposes. Central storage of DICOM content makes telemedicine practical. Incorporating DICOM content in a database enables customers to build electronic healthcare records applications, while using application development tools from Oracle or others.
A new Oracle Multimedia object type, ORDDicom, natively supports DICOM content produced by medical devices. This object type holds the DICOM content and extracted metadata, and implements the methods to manipulate the DICOM content. A new Java proxy class, OrdDicom, provides access to the ORDDicom database object through JDBC in a Java application. For applications that already store DICOM content directly in BLOBs or BFILES, a relational interface is provided as a PL/SQL package (ORD_DICOM).
By presenting DICOM content stored in a database as objects, Oracle enables both rapid application development and easy, secure management of large archives of DICOM content.
Oracle Database 10g Release 2 (10.2) provided support for extracting the most important metadata as DICOM attribute tags into an XML document, and then indexing and searching these tags to find DICOM content that matched certain conditions. Oracle Database 11g Release 1 (11.1) extends that capability by supporting complete and extensible metadata extraction. Customers can extract metadata according to an Oracle-specified XML schema, or create and use their own schema definition to extract subsets of the standard DICOM attribute tags or private tags. The extracted metadata can then be stored in a table to facilitate DICOM content searching based on standard or private DICOM attributes.
This enhanced metadata extraction capability enables customers to build large archives of DICOM content. By customizing extracted XML metadata documents, customers will be able to create highly specialized indexes to DICOM content based on standard and private DICOM attribute tags.
Oracle Multimedia can verify that DICOM content adheres to a set of user-specified conformance rules.
DICOM content is generated by many devices. While most conform to the DICOM standard, some do not. It is important to be able to identify DICOM content that does not conform to the standard, or to the conformance rules for a particular organization or enterprise. Validating DICOM content for conformance can ensure the consistency of a DICOM archive. It enables a database to accept DICOM content from multiple sources and verify the integrity of the content.
Oracle Database 11g Release 1 (11.1) adds methods and functions to copy and convert images from DICOM into other image formats (for example: JPEG, GIF, PNG, and TIFF), and to copy and generate scaled versions and thumbnail images. In addition, this release provides a set of optional methods and functions to copy and process (for example: compress, scale, rotate, and crop) image content, optionally during the conversion process.
To view medical images stored in the DICOM format in Web applications, a copy of the images must be created in formats that are compatible with the browsers that are currently used in the industry. Oracle Database 11g Release 1 (11.1) enables customers to automatically copy, reformat, and deliver DICOM images to applications that require popular industry-standard image formats such as JPEG.
Oracle Database 11g Release 1 (11.1) adds a method that makes new ORDDicom objects with DICOM content and extracted XML attributes anonymous, in accordance with the rules specified by an anonymity document. The anonymity document defines both the set of attributes that should be made anonymous and the actions to take to make them anonymous.
This method can be used to generate new, anonymous ORDDicom objects, assuring that users of a DICOM medical archive see only the DICOM content and metadata that they are authorized to see. For example, clinicians need full access to DICOM content and metadata for each patient they are treating. They must be able to view all the DICOM metadata included in DICOM content. Researchers, on the other hand, need only partial access to the same DICOM metadata for patients participating in a study. Patient privacy regulations require that this class of users not be permitted to view attributes and metadata included in ORDDicom objects that contain personally identifying information.
By providing anonymity services in the database, Oracle Database allows appropriate access for different classes of users of a DICOM medical archive, regardless of the application used to access the ORDDicom objects in the archive.
Oracle Database 11g Release 1 (11.1) includes the ability to generate new ORDDicom objects by combining digital images of various formats (for example: DICOM, JPEG, RAW, TIFF, and GIF) with an XML representation of the associated DICOM metadata. This operation results in well-formed and validated ORDDicom objects, which can be stored in a table in the database or delivered to a DICOM viewer. This feature is particularly useful for generating DICOM secondary capture images.
Storing and retrieving film-based medical images is expensive and prone to error. Clinical and research purposes require that some DICOM content be retained for extended periods of time. Replacing film-based medical images with DICOM images reduces storage and retrieval costs. Storing scanned images with their metadata in DICOM format can make non-DICOM images more useful. Using the same technique, new DICOM content can also be generated to correct metadata errors in the original DICOM content.
A key feature of DICOM support in Oracle Database 11g Release 1 (11.1) is that its run-time behavior is determined by a set of user-configurable documents. This set of documents is collectively managed by the data model repository. Administrators can update this data model repository to configure Oracle Multimedia DICOM for a particular database instance.
Hospitals need to be up and running at all times. They cannot shut down the system for any of the following reasons:
To update to a new version of the DICOM standard
To incorporate private DICOM attribute tags for a new piece of equipment
To change their DICOM conformance rules
To modify the set of DICOM attribute tags they extract from each ORDDicom object or to change the XML encoding of the extracted attributes
To modify their DICOM anonymity rules
This design enables customers to upgrade Oracle Multimedia DICOM at any time, without interfering with a running DICOM archive.