A Future Health Information Provider Architecture

• Abstract
• Introduction
• A categorisation of some HealthGrid projects
• A Wireless HealthGrid Project
• Opportunities for the HealthGrid in New Zealand?
• Acknowledgements
• References
• Footnotes

Abstract
The term "HealthGrid" identifies a cluster of biological, medical and health-related research and development projects formed in the European Union which utilise Grid computing to solve computing intensive tasks. It is expected that these HealthGrid projects will benefit health care service development and delivery within the decade. This paper provides a review of some of the worldwide initiatives in bringing Grid infrastructure and applications to the delivery of next generation electronic health care services. An example incorporating wireless networking and the Grid is presented and the paper concludes with a discussion on the potential direction for HealthGrid research and development in the New Zealand context.

Introduction
A HealthGrid can be described as an environment where data of medical/health interest can be stored, processed and made available to the appropriate actors within the health care system. The HealthGrid must enable secure storage and appropriate access to information based on actor roles such as physicians, nurses, health care centres and administrators as well as patients. Such a system respects the ethical requirements of the health care profession and the observance of regulations and related industry standards. So far this sounds like most health care information and communications technology (ICT) systems which are seen as large complex ecosystems with many complex information needs.

A HealthGrid can be further specified as a distributed^[a] computing and communication infrastructure that runs a middleware layer that enables the effective sharing/control of many computer, communication and storage resources and data repositories that can belong to a number of different actors, organised as one or more "Virtual Organisations" (VO).^[10] The combination of large numbers of computing, communication and storage resources co-operating across a number of organisations is known as a Grid - using the same concept as the power grid.^[11] An example is shown in figure 1 where a number of organisations may cooperate to provide a service incorporating information services or data acquisition in a cellular network scenario. In this example two wireless gateways offer access and services to cellular devices, these cooperate with a data depot, an external service gateway and a data federator, this is discussed further in section 3.

Figure 1: Grid concept - wireless information capture scenario

A VO is a group of organisations that jointly provide some set of services by supplying various computing, communication and information resources which, when combined, meet the objectives set out for the VO. These objectives may be business or personal goals and typically they are computationally dynamic and collaborative in nature. For example a VO may involve the set of patients, staff and organisations that form a radiology service. Such a VO can be a long-lived association being contracted to provide a service for a long period or short lived when providing an ad hoc service on demand. At the time a VO is established, members agree the terms and agreements (rules) that will be used to manage the VO, prescribing the degree of trust between the parties and specific rules for information exchange.

The resources in a computational grid are CPU cycles,^[b] data storage and communications bandwidth, which are allocated to processes and tasks under the control of resource management components. The computational Grid^[11] was originally envisaged by Ian Foster, Carl Kesselman and Steve Tuecke, who led efforts to create the Globus Toolkit,^[12] which is one of the major middleware^[c] solutions that provide control and application services to enable the Grid infrastructure. Globus incorporates CPU and storage management, security provisioning, data transfer management, infrastructure monitoring and a toolkit for developing services based on the Grid infrastructure. The Grid middleware services include agreement negotiation between actors in the Grid as well as notification mechanisms, trigger services and information aggregation services. The Globus toolkit^[9] is currently at version 4 (aka GT4) and its various versions have been used as the grid middleware layer for many e-science and collaborative applications. The task/process in a complex problem area can be managed by workflow engines^[d] and developed using, for example, Service Oriented Architecture (SOA) and Web services^[e] in order to provide automated and componentised solutions to these tasks.

Effectively, the combination of these computing and communication resources creates a virtual computer architecture used for solving complex or large scale problems, especially those that can be split up into parallel tasks. Initially, these were identified as scientific problems from areas such as the life sciences, medical research, pharmaceutical modelling, bioinformatics, etc. Examples include gene analysis, heart modelling, the simulation of drug affects on cancers, and molecular dynamics. Development of Grid applications for health care provision has formed as Grid technology has matured. One such health care project is e-DiaMoND which uses the Globus Toolkit version 3 (GT3) to implement a breast screening application. A blueprint of the framework used in the project has been published to aid development of related projects.^[5] The e-DiaMoND project was completed in November 2005 and has been "deemed to be of significant potential benefit to the NHS and to the female population of the United Kingdom".^[5] In the Virtual Organisations for Trials and Epidemiological Studies (VOTES) project, the development team used the Globus Toolkit version 4 in which they developed a fine-grained anonymisation service where identifying data in medical records is replaced such that transmitted data cannot be traced to a patient unless the users’ privileges are appropriate.^[22]

There are however a number of middleware solutions for the Grid as this is currently a major area of research. Among these middleware initiatives are several related to the "HealthGrid"^[14] such as the Enabling Grids for E-sciencE (EGEE),^[6] which is a European Union grid infrastructure project which has a health focus and which, for example, is being used as the infrastructure for studies on Avian flu. The EGEE project has developed a specific middleware called gLite^[8] which is gaining some momentum in HealthGrid areas because it is more lightweight than other Grid middleware to enable higher performance to be achieved. Projects such as "Health-e-Child", which is an integrated biomedical platform for Grid-Based paediatrics,^[13] have been developed on the EGEE infrastructure and the gLite middleware. Health Level 7 (HL7)^[f] has been identified as potentially crucial for developing the HealthGrid because of the widespread use of HL7 in existing healthcare applications and has been used for interaction with healthcare systems in projects such as e-DiaMoND, etc. The EU eHealth projects generally require the use of HL7 and these projects are listed in the ICT for Health: Resource Book for Health Projects report.^[7]

A HealthGrid depends on high speed networking. The integration of networking into the Grid is becoming more sophisticated with not only Internet Protocol connectivity but also very high-speed, optical circuit connectivity.^[20] In Europe, HealthGrid projects generally operate through the National Research and Education Networks (NRENs), which limits current integration with the health care system because most health care organisations are not connected to the NRENs.^[16]

As New Zealand’s advanced research and education network is due to be largely completed by the end of 2006, and as most New Zealand universities are developing grid computing centres, there exists an opportunity to investigate the use of some Grid based applications in the New Zealand context. This would begin to increase local competencies in these application of grid computing and furthermore identify the most promising overseas projects for exploitation in next generation healthcare delivery. This could allow the investigation of New Zealand specific health care issues and data analysis, mirroring the success of such studies in the Molecular Medicine Informatics Model (MMIM)^[g] in Australia where federated data has been used to lead the world in colon cancer therapy. With the health system taking increasingly more of the country’s GDP, we must intelligently provide a roadmap to encompass new technologies and optimise how ICT can aid in future health care service provision.

This paper provides a brief categorisation of HealthGrid projects in section 2. Then section 3 highlights a project integrating wireless devices for personal health services, which can be seen as an area of interest within New Zealand. In section 4 there is a brief discussion of how HealthGrid research and development could aid in New Zealand health care delivery.

A categorisation of some HealthGrid projects
There are a large number of health related Grid projects worldwide, with rapid progress being made in fields such as computer-based drug design, medical imaging and medical simulations, and there is a huge amount of data (clinical, genomic, proteomic, etc) in heterogeneous sources and formats. Medical informatics needs large computations including the analysis of 2D/3D/4D images, simulations (eg, for more effective radiotherapy delivery), etc. The computing resources for medical and health informatics must be accessible from medical centres and physicians offices, in order to deliver fast and reliable results.

The HealthGrid concept appeared in the literature in 2003, and there has been an international annual workshop from 2003 to 2006, which has provided a snapshot of largely EU related HealthGrid projects. These workshops are used as a basis for the analysis here. In 2005, the workshop was entitled "From Grid to HealthGrid"^[15] and was divided in four themes:

1. Knowledge and data management;
2. Deployments of grids in health;
3. Current projects; and
4. Ethical, legal, social and security issues.

The HealthGrid 2006 workshop covered "Challenges and Opportunities of HealthGrids"^[16] which grouped contributions into:

1. Medical imaging;
2. Bioinformatics;
3. Knowledge discovery;
4. Medical assessment; and
5. Ethical, legal and privacy issues.

Another perspective on the analysis of HealthGrids is to view them in terms of infrastructure, such as the technology employed in the EGEE project, the developed middleware such as gLite^[8] and the Globus toolkits, and end-user applications/projects, such as MammoGrid,^[19] e-DiaMoND, etc, which are typically the computing intensive tasks within healthcare delivery.

The key HealthGrid projects tend to fall under several broad headings:

Medical imaging: This area had the most conference contributions in 2006 and covered a wide area - neuroimaging, mammogram analysis, MRI, and their analysis and scheduling. There are opportunities in this area to provide valuable improvements to the New Zealand health care system, such as enhanced screening programmes. Some projects like e-DiaMoND have published their blueprints and have had a significant impact on health care. As a result, it is possible to replicate the infrastructure of such projects. Furthermore it is to be expected that some of these projects may be commercially exploited in the short to medium term, within circa two to six years. The financial impact of, for example, breast cancer in the US is estimated to be around US$6 billion^[h] so there is expected to be a significant commercial impact through provision of systems that can provide early and enhanced detection of breast cancer.

Bioinformatics: This was the second largest part of the 2006 HealthGrid workshop, covering analysis of genomics, protein structures, SARS and other infectious diseases. These projects seem to be of most use in longer term research rather than in real time or near real-time healthcare provision.

Knowledge discovery and medical assessment: This area is expected to have great application to health care systems in the near future. This area saw significant numbers of projects reporting progress at the 2005 workshop and a smaller contribution in 2006. This category includes projects like Health-e-Child^[13] for paediatrics and the ARTEMIS^[4] project for secure interoperable sharing of Electronic Health Records [EHR]. Medical assessment enhancements are being worked on in a number of projects such as projects working on enabling the sharing of medical expertise (through for example communities of practice), systems to provide paramedical aid, and improving effective resource management and patient outcomes, such as radiotherapy treatment analysis and planning. The development and deployment of such projects are likely to have a positive impact on health care delivery, patient outcomes and ICT systems and analysis of patient data. In general these initiatives look to be feasible for deployment in the medium term, circa three to six years.

Papers have discussed knowledge discovery and medical assessment, ethical, legal, security and social issues and grid infrastructure and application development. The key papers presented in the HealthGrid workshops in 2005 and 2006 are listed in table 1, classified into these areas.

Table 1: Key papers from the HealthGrid workshops in 2005 and 2006

The key issues facing HealthGrid developments that were raised at the 2005/6 workshops are:

• ethical, legal, security and social issues; and 
• the effective engineering of HealthGrid systems.

HealthGrid deployment scenarios are currently research-oriented and need to take a business focus for future development. HealthGrid application design patterns and components (such as for data anonymisation) are becoming clearer as the research projects mature, thus enabling the effective development of the next generation of grid middleware services, eg, such as the developments in the EGEE project.^{[6, 8]}

The HealthGrid Association^[i] is a community that gathers together individuals who are actively exploring the impact of HealthGrid technology on healthcare provision and research. The HealthGrid Association has identified three issues that must be addressed in order to develop this concept from today’s applied/industrial research focus through to deployment:

1. Availability of grid services (probably these will be web services based) - especially in the area of data and knowledge management.
2. Definition and adoption of international standards and interoperability across many levels:
   • Grid infrastructure and Grid middleware to provide robust, scalable and secure frameworks that can work with many different formatted data types and repositories;
   • Health applications, the move to HL7 version 3 may be of significance because of the move to an XML based message format which fits with the web services model used in Grid computing. HL7v3 has been used in its pre-standard form in a number of successful projects as has DICOM for medical imaging.
3. Deploying services in health care centres - the move from research and education needs to be managed through to deployment if proven to be effective in the health care system.

The EU SHARE project^[25] has the task of determining the roadmap for HealthGrid. Its findings and recommendations are due by the end of 2006 and were discussed within HealthGrid 2006^[16] but no further information was available at the time of writing this paper.

At this point, a generic roadmap of HealthGrid development could be envisaged as shown in figure 2. We are about three years into many of the first set of projects and the key issues to work through are electronic health records (EHR), standards, security and interoperability. It is likely that the first deployments will be in the area of information and data services, identified as area^[I] in figure 2 - especially within the EU where they are expected to support EU eHealth initiatives.^[7] A second group of projects identified as^[II] in figure 2 cover, for example, decision support systems, personal health systems, including wearable health sensors, wireless health services and biomedical informatics are also underway, however these are generally expected to be deployable in a second phase somewhat later than information and data service based HealthGrids.

Figure 2: HealthGrid development and deployment time scales 

A Wireless HealthGrid Project
One of the areas being developed in the use of Grid computing for health care provision is an integration with cellular and wireless technology within a mobility context. The scenarios envisaged for the use of Grid and cellular technology include the enhanced provision of emergency services and ad hoc delivery of services for chronic diseases. A number of process models relating to "information on demand" and information management are envisaged such as: ECG/blood pressure/pulse transmission; and information flow/transformation within and between healthcare providers. Such sensor networks can provide the players in a virtual organisation - physicians, nurses, patients - with individual and current information anytime, anywhere. Provision of information or alerts using mobile phone technology such as the short message service (SMS) have given positive health outcomes in various small scale studies. These can be developed to large scale deployment through the use of Grid computing resources. Mobile surgical and intensive care systems could make use of on-demand computing and communication services, such as access to distributed knowledge sources and the application of distributed calculations; one example that has been suggested is the "on demand" pattern analysis of radiological images.

Future wireless/cellular devices may be able to transmit vital sign information, enabling remote patient monitoring. Such applications may require the provision of large amounts of processing power and the development of a number of virtual organisations that consist of many players in the healthcare system. Integration of wireless networks such as third generation (3G) and beyond 3G (B3G) or 4G wireless systems with Grid computing can be a solution, especially in the area of sensor Grids that collect large data sets from remote sensors. This section overviews some of the work being undertaken in this area which may be of potential benefits for New Zealand in the medium term, especially for a remote heart monitoring and future emergency provider scenario.

One of the moves within the Grid computing fraternity has been the migration from traditional high performance and distributed computing to pervasive and utility computing based on the advanced capabilities of the wireless networks and lightweight, thin devices. These are areas being worked on at Victoria University of Wellington (VUW) based on the creation of VOs for wireless service provision.^[17] There are several wireless initiatives in New Zealand^[18] and a number of system vendors and service providers interested in developing services for the health care area that may benefit from the integration of Grid computing with wireless/cellular device capabilities.

The Akogrimo project,^[2] consisting of 14 EU organisations, which is one of the leading projects in the wireless/Grid area and is developing the "Mobile Grid" to illuminate the potential synergies between todays 3G wireless systems and Grid computing power. The key eHealth project within Akogrimo is the creation of a heart monitoring and emergency provider scenario.^[3] This work is seen as primary research and development in the application of 3G and Grid to eHealth and the vision is illustrated in figure 3 which is taken from the Akogrimo web site.^[2] The Akogrimo validation scenarios cover eHealth and disaster management; integrating Grid services architectures in mobility scenarios using cellular/wireless networking based on knowledge oriented applications.

Figure 3: Akogrimo vision^[2]

The key players in the Akogrimo heart monitoring scenario are shown in figure 4. Patients can be monitored through sensors they wear which are integrated with cellular devices and wireless devices in vehicles that feed data to information warehouses for processing. The VO is made up of network providers, emergency services and Grid infrastructure providers. The scenario is further discussed in the white paper from Akogrimo.^[3] In figure 1, a number of components are identified, such as wireless gateways, service gateways, data depots or warehouses and information federators that make up such a Virtual Organisation.

Figure 4: Akogrimo health monitoring application, taken from reference^[3]

Other projects are also investigating similar scenarios; one such is SIGNAL^[23] whose architecture is shown in figure 5. This type of three layer architecture provides a generic perspective on wireless Grid systems based on SOAP^[j] for transmission over HTTP and SOA using web services to provide a loosely coupled solution to wireless information system development. Integration with location determination technology would be required, such as the current development in mobile phone cellular systems which are required in many countries to determine a customer’s location for emergency service provision. This is enabled through the integration of GPS within the handset or through triangulation mechanisms derived from signal propagation with cell towers. Interaction between wireless/cellular devices and Grid computing is enabled through a proxy gateway that is enabled dynamically to provide privacy and quality of service. Additionally caching features provide service/transmission efficiency. Such systems can employ Geographic Information Services (GIS) to map the location of a customer, eg, through OpenGIS (http://www.opengis.org/) which enable access to mapping services based on web services which enables easy integration with Grid computing services.

Figure 5: Signal architecture, taken from^[23]

Opportunities for the HealthGrid in New Zealand?
The projects identified in section 3 emphasise the scope of HealthGrid for real time monitoring and emergency service provision through an integration of healthcare providers, computing and communication service providers. The scope of HealthGrid is potentially huge - from pure research-oriented projects, to projects providing medical knowledge dissemination, to near real time services such as image analysis, right through to vital sign monitoring and emergency service provision. These projects are enabled through high speed networking and Grid computing technology.

With the development of the New Zealand research and education network^[1] and the developing Grid capabilities within New Zealand universities, New Zealand is on the cusp of better deployment of Grid technology toward the development of eHealth solutions. This Grid resource, enabled through our next generation Internet capability, can provide a good base line for experimenting in tomorrow’s eHealth infrastructure. This opportunity to explore the use of Grid technology and applications within health care should be grasped firmly within the next few years in order to determine the benefit to New Zealand health care delivery and to further develop the local ICT industry.

From a HealthGrid deployment perspective it could be feasible to envisage New Zealand operating within a global or International HealthGrid (IHG) which would provide access to international health initiatives and services. A National HealthGrid (NHG) for effective local Grid-based service and application delivery may soon be within reach. This national focus would include health centres and ICT providers and universities. It could be envisaged that national HealthGrid initiatives may be targeted at key local activities, as shown in figure 6. This might include the equivalent of automated mammogram analysis, child health monitoring as envisaged in Health-e-Child, and cellular sensor networks to enable distributed patient monitoring. International linkages could be envisaged for the global sharing of medical expertise, for addressing global health challenges, etc.

Figure 6: HealthGrid New Zealand?

What are the HealthGrid opportunities? What are the time scales for HealthGrid - internationally and nationally? What activities are required to move New Zealand forward in the HealthGrid? Some of these questions may be answered with the outputs from the EU Share project^[25], although applicability to the New Zealand context needs to be tested. The HealthGrid association aims to make its knowledge base available in December 2006. This is the time to further develop New Zealand expertise, our roadmap and our national and international collaborations towards enabling the HealthGrid.

How could this be achieved? A number of initiatives could be envisaged in the short term to kick start this process:

• Develop business models to identify the key HealthGrid projects that will most benefit New Zealand. This requires an understanding of likely deployment models and the virtual organisations required nationally and internationally.
• Engage existing New Zealand expertise in Grid technology and medical/health research. 
  • for example, further enable international collaboration to learn from the key HealthGrid projects and develop relationships with potential HealthGrid vendors such as IBM. 
• Connecting the next generation health network to the advanced research and education network would enable the initial development of research based Grid applications,
  • for example, it may be feasible to emulate the positive experience of the Molecular Medicine Informatics Model (MMIM) project^[26] within New Zealand.
• To enable the health ICT vendor community in New Zealand we should build capability in HealthGrid technologies in order to provide an appropriate knowledge base. This capability building requires:
  • Local test beds for healthcare providers and vendors to work with;
  • Local experts with a knowledge of HealthGrid standards and technologies; and
  • Engagement with international research.

Within HealthGrid, the next few years will certainly produce some interesting outputs that will begin to clarify how New Zealand could grasp this key new technology in tomorrow’s health care provision.

Acknowledgements
Thank go to Victoria University of Wellington for funding this initial review through the University Research Fund for 2006. Thanks to the reviewers for their comments and feedback.

References

1. Jarvie C. The advanced network background. MoRST. http://www.morst.govt.nz/uploadedfiles/Documents/Current%20work/ANbackground.pdf. Accessed 26 April 2006.
2. Akogrimo Project. http://www.akogrimo.org/. Accessed 26 April 2006.
3. Loos C. E-Health with mobile grids: the Akogrimo heart monitoring and emergency scenario. Akogrima Project, 7 February 2006. http://www.akogrimo.org/download/White_Papers_and_Publications/Akogrimo_eHealth_white_paper_short_20060207.pdf. Accessed 26 April 2006.
4. Artemis. http://www.srdc.metu.edu.tr/webpage/projects/artemis/. Accessed June 2006.
5. Brady M, Gavaghan D, Harris S, Jirotka M, Knox A, Lloyd S, Politou E, Power D, Simpson A, Slaymaker M, Williams J. e-DiaMoND: the blueprint document, 4 November 2005.
6. http://www.ediamond.ox.ac.uk/publications/blueprint-Final.pdf. Accessed 17 June 2006. Enabling grids for E-sciencE. http://www.eu-egee.org/. Accessed 17 June 2006.
7. ICT for health. Resource book of eHealth Projects, FP6, European Commission. http://www.srdc.metu.edu.tr/stakeholders_group/public_docs/fp6booklet%20Patient%20identity.pdf. Accessed 17 June 2006.
8. gLite: lightweight middleware for grid computing. http://glite.web.cern.ch/glite/. Accessed 17 June 2006.
9. The globus alliance. http://www.globus.org/. Accessed 26 April 2006.
10. Foster I, Kesselman C, Tuecke S. The anatomy of the grid: enabling scalable virtual organizations. Int J Supercomput Appl 2001 15(3).
11. http://www.globus.org/research/papers/anatomy.pdf. Accessed 4 September 2006. Foster I, Kesselman C. Computational grids. In The grid: blueprint for a new computing infrastructure.. Morgan-Kaufman; 1999.
12. Foster I. Globus toolkit version 4: software for service-oriented systems. IFIP International Conference on Network and Parallel Computing. Springer-Verlag, LNCS 3779; 2005: 2-13.
13. Health-e-Child, http://www.health-e-child.org/. Accessed 17 June 2006.
14. Health grid association, http://www.healthgrid.org/. Accessed 26 April 2006.
15. From grid to healthgrid, Proceedings of HealthGrid 2005. In Studies in Health Technology and Informatics March 2005; 112. Solomonides T, McClatchey R, Breton V, Legré Y and Nørager S, eds.
16. Challenges and opportunities of HealthGrids. In Proceedings of HealthGrid 2006. Hernandez V, Blanquer I, Solominides T, Breton V, Legré Y, eds. IOS Press; 2006.
17. Komisarczuk P, Bubendorfer K, Chard K. Enabling virtual organisations in mobile networks. IEE 3G2004 Conference, October 2004, London, England.
18. Komisarczuk P, Webster B, Welch I, Cullen R, Duke J, Wright T. Developing a framework for nursing informatics - a NZ initiative based around district health nurse mobility. HINZ 2005. Auckland: New Zealand; August 2005.
19. MammoGrid. http://www.mammogrid.com/. Accessed 26 April 2006.
20. Minoli D. A networking approach to grid computing. Wiley; 2005.
21. Bilykh I, Bychkov Y, Dahlem D, Jahnke JH, Kuziemsky C, McCallum G, Obry C, Onabajo A. Can grid services provide answers to the challenges of national health information sharing? Proceedings of the 2003 conference of the Centre for Advanced Studies on Collaborative Research. Table of contents. Toronto, Ontario, Canada: 39-53
22. Stell AJ. Sinnott RO, Ajayi O. Secure federated data retrieval in clinical trials. TeleHealth 2006 Conference, Banff, Canada, June 2006.
23. Scalable inter-grid network adaptation layers (SIGNAL). http://www.wirelessgrids.net/. Accessed June 2006.
24. Breton V, blanquer I, Hernandez V, Legré Y, Solomondes T. Proposing a roadmap for healthgrids. HealthGrid 2006, Valencia, Spain, 7-9 June 2006.
25. Supporting and structuring healthgrid activities & research in Europe: developing a roadmap. http://www.eu-share.org/. Accessed June 2006.
26. Molecular Medicine Informatics Model, http://mmim.ssg.org.au/whatis.htm, last visited August 2006.
    

Footnotes

1. a Distributed computing refers to computing systems in which services to users are provided by teams of computers collaborating over a network. In a distributed system tasks are divided among several computers rather than having all processes originating from one central computer. Client/server systems are one type of distributed computing system.
2. b CPU - Central Processing Unit. "CPU cycles" execute computer instructions
3. c Middleware: software providing communication, conversion/translation or services in a distributed system. d The Workflow Management Coalition (http://www.gridworkflow.org/snips/gridworkflow/space/WfMC) define workflow as "The automation of a business process, in whole or part, during which documents, information or tasks are passed from one participant to another for action, according to a set of procedural rules". In the context of Grid computing the term workflow usually concerns the automation of distributed IT processes and is enabled through a computational engine enabled through a workflow enactment service.
4. e A Web Service is a software component that is described via the Web Services Description Language (WSDL) which uses an XML message format to access the service via standard network protocols such as SOAP over HTTP.
5. f HL7 is an ANSI-accredited standard defining clinical and administrative data and messages in the healthcare industry.
6. g MMIM: http://mmim.ssg.org.au/whatis.htm
7. h http://www.ngi-nz.co.nz/applications/x-ray.html
8. i HealthGrid Association - http://www.healthgrid.org/.
9. j SOAP - Simple Object Access Protocol, HTTP - Hypertext Transfer Protocol, SOA - Service Oriented Architecture