HealthCyberMap - PhD Research Project by Maged N Kamel Boulos

HealthCyberMap: Mapping the Health Cyberspace Using Hypermedia GIS and Clinical Codes

PhD Research Project by Maged N Kamel Boulos
(PhD awarded in December 2002)
Centre for Measurement and Information in Medicine, School of Informatics, City University, London, UK

Project Summary

OBJECTIVES AND DESCRIPTION

Download the Most Up-to-date Project Summary, Aims and Objectives (PDF)

HealthCyberMap aims at mapping the health information cyberspace in unique and novel ways to help consumers and providers of health information navigate and plan this complex "virtual" space more efficiently and effectively. This is achieved through intelligent categorisation and interactive hypermedia visualisation of the health information cyberspace. Customisation possibilities are included in the study, e.g., based on the user's geographical location to deal with language as well as any specific health needs/ online resources related to this location. HealthCyberMap uses "conventional" geographical maps to map cyberspace, attributing health resources on the Web to the geographical location (country) of their corresponding health information providers. Another set of Human Body topic maps is used to navigate health resources by category/ topic/ body system according to the WHO International Classification of Diseases (ICD) which acts as HealthCyberMap's medical ontology. We also use a third type of cybermaps, our Virtual Health Library map, to structure and organise health information resources by type: electronic articles, electronic journals, e-books, digital atlases (e.g., of dermatology), audio-visual material, other online health services, etc. HealthCyberMap hypermedia zoomable maps are used to locate and launch health resources on the Web, and also to display their attributes (bibliographic card). The study also demonstrates the feasibility of Electronic Patient Record to HealthCyberMap problem to knowledge linking using ICD-9-CM codes as crisp problem to knowledge linkers or knowledge hooks.

TOOLS AND METHODS

HealthCyberMap is a Web-based project developed using ESRI ArcView GIS v3.1 with BodyViewer (ICD-9) and Internet mapping extensions (WebView 1.1). ASP server pages are used to query HealthCyberMap database. The pilot version of the project only maps a limited number of health information providers world-wide and just part of their resources. Aided by conventional meta-search Web engines/ tools and existing medical directories like OMNI, candidate resources are hand-selected, and their attributes, including ICD codes representing their subjects as well as any recognised quality rating they already bear, are manually compiled in HealthCyberMap database based on the Dublin Core metadata set. ICD-9-CM code locators (e.g., Yaki Technologies' eICD.com) are used to locate the ICD codes that best describe the topics covered by a given resource. (ICD-9-CM codes are maintained by the American Public Health Service and the Health Care Financing Administration of the US.)
See HealthCyberMap project in ArcView.

RESULTS

The first public pilot site release of HealthCyberMap (URI: https://www.healthcybermap.org or https://healthcybermap.semanticweb.org) is now online. A small-scale public evaluation of HealthCyberMap concepts (questionnaire) will be conducted during 2002 and results analysed and published.

HealthCyberMap adopts a model-based metadata framework that builds upon one or more clinical coding schemes for the semantic classification, visualisation and browsing of medical information resources on the Internet. HealthCyberMap utilises GIS methods to generate its navigational cybermaps, which are also based on sound cartographic principles and familiar medical metaphors. This maximises the spatial legibility and utility of the generated hypermaps. HealthCyberMap's metadata base has been modelled in Prot g -2000 and imported to ArcView GIS for mapping and cyberspatial analysis (detecting infogaps using choropleth maps of the human body). HealthCyberMap's cybermaps can be considered as spatialised browsing views of its underlying metadata base. Literature covering the bibliographic and cybergraphic uses of GIS is very rare (and none is specific to the medical domain), and so is research into the relatively new methods of spatialisation for browsing Web resources. Using a clinical coding scheme as a metric for spatialisation (semantic distance) is unique to HealthCyberMap. Also HealthCyberMap's use of BodyViewer (an ArcView extension) is novel and unconventional.

CONCLUSIONS

HealthCyberMap addresses many of the current knowledge needs of Internet health information providers, healthcare professionals and the general public and patients, by helping them better understand, plan and manage the health information cyberspace. Problem-specific knowledge (the right, contextually relevant knowledge) linked to real patient data is the key to informed clinical decision making and better healthcare outcomes.

ADDITIONAL NOTES

1. The geographical cybermaps are special maps where distance, direction and time loose their conventional meanings (e.g., you can have two resources from two different countries displayed at the same time in two Web browser windows, but you cannot physically visit two countries at the same time).

2. The current HealthCyberMap Bibliographic Card (Resource Metadata) includes the following fields:

HealthCyberMap Bibliographic Card
(dc=Dublin Core; hcm=HealthCyberMap-specific extensions)

dc:Creator="Author(s) name(s)"

dc:Title="Resource title"

dc:Subject (1, 2, 3)="ICD Code, e.g., E948.2"

dc:Description="Textual equivalent (description) of ICD code,

e.g., CHOLERA VACCINE"

dc:Publisher="e.g., WHO"

dc:Date="Resource date of last update"

dc:Type="The category of the resource, e.g., electronic article,

         fact sheet, electronic journal paper, collection, e-book,
         digital atlas, audio-visual material, interactive
         resource, event, software, other online health service"

dc:Identifier="Resource URI"

dc:Language="e.g., en"

dc:Coverage="The spatial extent or scope of the content of the

resource in the form of a spatial location (a place

name or geographical coordinates from the Thesaurus

             of Geographic Names)"

hcm:Location (City, Country)="Publisher or author(s) geographical
                                                 location, whichever is more
                              relevant"

hcm:Quality="Level of evidence (official guideline, systematic

review, RCT, other peer-reviewed studies, official

CAT, expert opinion), recognised code of ethics

or quality seal, Trusted Publisher or Listed in
Trusted Directory, e.g., OMNI, or unspecified"

hcm:Comment="Any additional information"

N.B.: We don't distinguish between public, patient and professional audiences on purpose (the boundaries are very hazy in this open world of free information and patient-centred care/ self-help/ self-care). The three audiences are partners in healthcare.

3. Future directions include:

Online mapping of real world health problems (in addition to mapping the health information cyberspace), mapping the growth and distribution of Internet users and health-related Web sites, and a "click-thru" geographical visitor analysis and reporting service (cyberspatial analysis). This should help providers of health information on the Web, and healthcare managers in general, plan better services (including health education programmes) by understanding their users' needs and what competitors are offering. We are also planning to integrate services like a healthcare facility finder/ router and a travellers' healthcare advice utility.
Adding procedural codes and allowing users to search ICD-9-CM (or any other coding scheme we might use) from within HealthCyberMap using natural medical language (allowing for synonyms and alternative descriptions and also making it easier to find related entries like parent categories and subtypes). Instead of displaying the retrieved clinical codes, the search engine would retrieve the corresponding HealthCyberMap resource records (coded using these codes). The [ICD-9-CM codes/ descriptions] and [HealthCyberMap] tables of HealthCyberMap database will be linked based on common clinical code fields.
Combining different search/ visualisation criteria, e.g., show only resources on Tuberculosis on HealthCyberMap World Maps (vs. all resources), or link the EPR to only Guidelines on the relevant topic (vs. all types of resources).
Investigating ways and tools for ensuring HealthCyberMap database is always up-to-date (e.g., including a rot or broken link checker) and compatible with related initiatives and standards like Topic Maps (ISO) and Semantic Web (W3C).

Project References

A partial list is available in a separate 'Webliography/ Bibliography' section on this server.

Publications

A list of publications arising out of this project.

Contact Information

Maged N. Kamel Boulos
M.B., B.Ch., M.Sc. (Derm. & Vener.), M.Sc. (Medical Informatics), Ph.D. (awarded in Dec. 2002)
(I now work as Lecturer in Healthcare Informatics at the School for Health, University of Bath.)
Formerly: Research Assistant in Medical Informatics and Ph.D. Candidate
Centre for Measurement and Information in Medicine
School of Informatics
City University
Northampton Square
London
EC1V 0HB
UK
Fax (+44) 08709157698
E-mail: mkamel@medicad.com