S 4.3 Making DICOM work in the Netherlands Using Minimal datasets and Application profiles

M.G.J.M. Gerritsen, M.H. Baljon, J.A.T.G. van Kester, F.M. van Rappard, A. Venema.

Interuniversity Cardiology Institute of the Netherlands
PO box 19258, 3501 DG Utrecht, the Netherlands.

Summary

A DICOM standardization initiative has been started in 8 university cardiology centers in the Netherlands. These sites will define and exchange DICOM data sets. The project is the second phase of IMFOMED2 (IMages For MEDicine phase 2). Its objectives are:
• the specification of DICOM minimal datasets,
• the validation of DICOM datasets supplied by vendors and
• the exchange of DICOM messages.

  The project is mainly user-driven, which places it in a unique position to bridge the information gap between vendors and users. The expectation seems justified that IMFOMED will help DICOM to become a standard that will garantee image exchangeability and will contribute to realisation of cost-effective quality improvement in cardiology.

A project was initiated in the Netherlands for the exchange of cardiac image data between university centres in 1992. This project, known as IMFOMED ( IMages FOr MEDicine.) is coordinated by ICIN. ICIN is an institute for combined cardiology research in which all 8 Dutch university cardiology centres take part. As DICOM 3.0 is the standard in Medical Imaging [1], IMFOMED has put a first priority on the DICOM standardization in the participating cardiology centres.

The IMFOMED project is supported by a grant of the dutch Ministry of Economic Affairs. IBM and OSIX BV are partners for hard- and software support and advice.

A prototype of a DICOM viewer was developed during the first phase of IMFOMED. This application retrieved image data that were stored in DICOM format and related data that were stored in a relational database[2].

The project will generate and implement applications for the validation and exchange of DICOM messages among cardiology centres during the second phase. The applications, as described in more detail below, will consist of three main parts:

1. defining minimal data set requirements for clinical research studies,
2. validation of DICOM messages and DICOM sets stored on CD-R and
3. the transfer of DICOM messages via the Internet (prototype).

   The DICOM standardization process and the role that can be played by a user platform will also be discussed.

figure1: part of the DICOM information model

A DICOM information model is depicted in figure 1, which shows the relationships between a number of so-called 'information entities', such as Patient, Study and Series. Each entity has a number of attributes. A minimal data set (MDS) specifies which entitities and which attributes are mandatory. The major part of the attributes contain image related data like names and unique numbers for identifying the patient, the study, the medical staff invlved, the parameters of the device that generated the images and many other data .
The necessity of the presence of certain entities, modules and attributes will depend on the purpose of the information exchange. For example the exchange of image data between cardiology centres for research purposes will require data different from those needed for patient care or telemedicine. Therefore, each type of interchange demands its own "application profile". A separate MDS requirement can so be defined for each IMFOMED application. An application designed for this should store these data set specifications in an easily updatable relational database management system with flexible retrieval and reporting features. We think that using an interface like this establish a consensus between cardiologists on the MDS's to be used in daily practice.

The validation of DICOM data-sets stored on CD-ROM or transferred via Internet should be a basic component of any DICOM application. The IMFOMED project will design a DICOM checker to ensure conformity of the DICOM data set to the MDS requirements. As described above the MDS will vary with the application profile that reflects the purpose of the information exchange. In this way, checks can be specified according to the needs of each situation. This feature makes IMFOMED's DICOM checker rather different from those already available on the Internet today.

Use of the information super-highway for exchange of cardiologic images and related information is a very exciting development. It could become the way to gain independency of computer soft- and hardware platforms. It is also an area of great concern due to the privacy control implications. The easy access to the Internet requires extra security measures to be taken before data can be safely transferred via Internet or any other public network. The technical feasibility of using the Internet for the transfer of cardiologic images will be investigated during the IMFOMED project. The Integration of IMFOMED's DICOM checker in the Internet communication set-up will be one of the most important features.

The most critical factor for the success of the IMFOMED project is the increase of user involvement in the DICOM standardization process. Making tools for the definition, validation and exchange of DICOM data sets available to user-groups will be one of the main strategies to achieve that specific aim. Reactions received from the user community after the introduction of IMFOMED's MDS application already seem to justify the expectation that our approach will be effective in this respect.

Another important aim is to offer an organised user platform to vendors of DICOM solutions that will allow the development of DICOM applications which conform to the requirements of the user community. Reaction received from vendors have thus far been very positive in this respect[3].

Other areas of attention are the integration of the DICOM standard with other medical communication standards such as HL7 and EDIFACT. Communication servers that provide this type of integration are a very promising development in this area and will be closely monitored by the project-team and the participating centres[4,5].

Clearly, the IMFOMED-applications described above are not viewed by the participating centres as the ultimate goal of the project. They are intended to form the foundation of a successful implementation of the DICOM standard in Holland and throughout the rest of Europe, that will exploit its potential to produce cost-effective solutions and quality improvements on behalf of patient care and cardiology research.

[1]	Robert Hindel, PhD(ed.); Implementation of the DICOM 3.0 Standard, A Pragmatic Handbook, 1994 by the Radiological Society of North America, 2021 Spring Road, Suite 600, Oak Brook, IL 60521.
[2]	IMFOMED (IMages FOr MEDicine) W.A. Dijk, M.G.M. Gerritsen, INCIS International Workshop Conference, april 5-6, 1994 Mainz.
[3]	Holland Goes DICOM, M.H. Baljon, M. Gerritsen, Proceedings Computers in Cardiology, September 10-13, 1995 Vienna, IEEE Computer Society Press, Los Alamitos, California.
[4]	Linking PACS and HIS, December 1994, Mitra Imaging Inc., 115 Randall Drive, Waterloo, Ontario N2V 1C5 Canada.
[5]	Dejarnette and UMAB Develop DICOM/HL-7 Interface Gateway in: PACSnews pg 3 vol.3 issue1, November 1, 1994.