1. Introduction

This document is intended to provoke an optimized mindset for the development of a coherent technological policy and process which will guide the development of the next generation of intelligent telemedical communication systems. Optimal perception of what is both plausible and necessary requires vision; a new vision which aims to integrate the relevant features of emerging technologies with the well established need of providing quality medical services across space and time. Such a vision requires insight into both the technological and practical aspects of telemedicine. The conceptual basis for beginning the process of designing, implementing and refining the next generation of distributed telemedical systems is presented here. The Med-Wide-Web, a secure medical communication infrastructure utilizing Web compliant interface and communication protocols, will be the "infrastructure of discourse" for the following discussion. The Med Wide Web (MWW) poses the potential for a philosophical and conceptual recasting of that fundamental notion of telemedicine.

Web-based medical communication infrastructures distinguish MWW from historical telemedical processes yielding the next generation of the art: the Distributed Medical Collaborative Network. This paper attempts to offer the latest (Mar 97) thoughts and technological reevaluations which have gone into redefining the notion of health care 'at a distance.' This redefinition extends a communication bridge across administrative and specialist borders (i.e., intersite/intrasite). Through the cooperative workings of many, the goal is to bring to the table a new plan for Distributed Medical Intelligence and services. The objective of this recommendation, then, is to provide guidance for the future development and implementation of new technologies which will transform and extend telemedical practices in intelligent and empowering ways.

1.1 Participants

• Telemedicine Technologies and Support (ECU)
• Distributive and Collaboratory Environments (NPAC)
• Interface Design and Systemic Integration (I3)

1.2 Key Terms:

Distributed Medical Intelligence; Care Portal; Docking Station; Bridge; Cooperative Care; Web-based Distributive Collaboratory Environments; Medical Extranets; Session tracking; Machine Resident Intelligence; User Tailored Interface & Content; Computer Supported Cooperative Work; Cybercology

2. Current Telemedicine: Status and Limitations

NASA has defined telemedicine as "the integration of telecommunications technologies, information technologies, human-machine interface technologies, and medical care technologies for the purpose of enhancing health care delivery (Federal Telemedicine Gateway Menu: NASA)." In satisfaction of these criteria, East Carolina University developed and presently operates one of the most sophisticated and active telemedical infrastructures in the country. Between August 6, 1992 and February 10, 1997 ECU Medical School successfully conducted 1329 medical consults in virtually all areas of clinical medicine and surgery. ECU's work has more than proven the value and basic efficacy of telemedicine's capacity for delivering quality care at a distance, both in real time and in a store and forward mode.

Concurrent with this work, communication technology continues its incessant progress with better video teleconferencing, image and sound re-production, diagnostic imaging, medical record data provision, and so on. Over the last several years, however, Dave Balch, Director of ECU's Telemedicine program and Dave Warner MD, NPAC Nason Fellow and Director of the Institute for Interventional Informatics (I3), began to discern severe limitations in main stream models of telemedicine practice. To preface, there are no necessarily wrong applications or badly misguided uses of the applications thus far. Rather, availability and use of new technologies create suspicions about better technologies and still better uses of those technologies posing serious challenges to current telemedicine. This is the case for both the World Wide Web and languaging tools such as Java. These tools create wholly new conceptions of practicing medicine 'over the wire.'

Telemedicine, as it is practiced now, operates primarily on a point to point basis. One medical expert treats one patient at a time, or provides consult services to one remote physician at a time. Historically, interaction with a patient by more than one remote expert has been systemically precluded. Moreover, recording and filtering of exchange elements (content) has been largely incidental. Video and/or audio tapes are what is left with which to manually and laboriously collect, analyze, derive generically (hopefully) useful content and archive the rest. This is practically historiography or info archaeology-not an optimal use of medical human resources. Finally, most current infrastructures of telemedicine do not provide support for self optimizing interconnection of caregivers within the same facility. A cardiologist is unable to utilize the informatic infrastructure to interact with a pathologist in the same clinic or hospital. Given the growing salience of these and other impediments to distance practitioners, the time is ripe to propose an altogether different kind of integrative medical communications infrastructure. The first iteration of such an infrastructure was proposed in the Fall of '95 as Distributed Medical Intelligence: A Systems Approach for Developing an Integrative Healthcare Information Distribution Infrastructure (Warner et al., 1995). This work was then presented in January, 1996 at the Medicine Meets Virtual Reality Conference #4 in San Diego California.

3. Distributed Medical Intelligence: Telemedicine Wakes up to the Web

Distributed Medical Intelligence focuses on the process of communicating medical knowledge and service, not just the technological and administrative aspects of developing a medical informatic infrastructure. The main conceptual point is that the ways in which we think about technology determines which technologies are used thus restricting what we may/will use them for. Philosopher of science, Thomas Kuhn had ideas which apply very well to the present-at-hand fork in the road confronting telemedical culture:

a new theory, however special its range of application, is seldom or never just an increment to what is already known. Its assimilation requires the reconstruction of prior theory and the re-evaluation of prior fact, an intrinsically revolutionary process…Normal science does not aim at novelties of fact or theory and, when successful, finds none. New and unsuspected phenomena are, however, repeatedly uncovered by scientific research, and radical new theories have again and again been invented by scientists. History even suggests that the scientific enterprise has developed a uniquely powerful technique for producing surprise s of this sort. If this characteristic of science is to be reconciled with what has already been said, then researcch under a paradigm must be a particularly effective way of inducing paradigm change. That is what fundamental novelties of fact and theory do (Kuhn, 1962; 7,52).

Given a certain set of premises, certain kinds of perceptions, procedures and data follow accordingly, necessarily. If development of communication systems for medicine at a distance is simply going to amount to conceptualizing healthcare in more or less traditional ways with new kinds of electronifications and sensory equipment, particularly from the perspective of medical knowledge production, integration and implementation through practice, then all the rest of what is possible remains invisible to all involved.

The thinking which formed the conceptual basis for DMI occurred when experiences gained within the environment of clinical and experimental medicine were combined with experiences gained from participating in the early and very dynamic phases of the global information culture. Particularly influential was involvement in the human-computer interface, graphical visualization and distributed supercomputing explosion of the late 80's and early 90's. Exposure to such technologies as virtual reality, multimodal interfaces and the World Wide Web, provided an empowering perspective and some fundamental insights as to the transformational nature of these information technologies. The simultaneous experiences of being professionally oriented towards traditional clinical medicine and being an active participant on the bleeding edge of this technological wave forged a unique mindset. This mindset led, in part, to a new vision of telemedicine as a perfect and timely application for the Web.

As health care professionals, we are in an historically opportune position to more fully grok (i.e., actively seek to fully comprehend) and utilize the enhanced capacities for enriched communicative interaction that the Web provides. The MWW enables us with greater accessibility to a wider range of knowledge tools and communication resources. This, in turn, extends our telemedical capacity to deliver quality, care improving services to a wider range of persons in need. Conceptual, pragmatic and techno-procedural foundations of medicine itself will undergo subtle and great revolution if seminated by the principles of ubiquitous extensibility and need specific inter-connectivity.


3.1 DMI: Functional Elements

The basic structural features of an intelligent distributed medical communications infrastructure are as follows:

Components:

1. Point of Need (Care Portal): Any site which requires an informatic intervention. Optimally, a remote and medically specific environment instrumented with an operational tool kit of diagnostic and evaluative technologies for diverse and quantitative assessments of patient biological states. These devices connect to the local processor (s) in the care portal and then to the MWW via a central processor known as the Bridge (see below). The care portal is also designed to provide educational content to patient and medical personnel for both immediate and more long term purposes.

2. Point of Expertise (Docking Station): a 'lookout' of sorts from which a medical expert gives consultation, collaborative input and education. Highly specialized interfaces designed to optimize an expert's ability to provide the highest quality of service characterize the docking station's environment. Docking stations also connect to local processors and then on into the MWW via bridge mediated access. Docking station personnel will perceive and interact with a great diversity of medically relevant information coming in firstly from the care portals and secondly from multiple [other] experts participating in the same exchange. This exemplifies the value and power of the Web-based collaborative matrix vs. the traditional telemedical point to point connections! Processual and technological aims are to empower an expert so as to optimize their capacity for intelligently, competently and effectively responding to the point of need.

3.Primary Communication Server (Bridge): an intelligent medical communication hub which optimizes the flow of information between the care portals and the docking stations. The bridge is the first point of contact for the care portal. It serves as a request refinery directing the care portal to the most relevant resources to service their needs, whether a knowledge base or a live expert. The bridge is the central nervous system of the MWW organism. Bridge function maintains dynamic connectivity, data tracking, and data base access. It proactively draws, supplements and anticipates patient and expert inputs/outputs, provides the ancillary information of patient and expert education, and filters, refines and directs knowledge. The bridge optimizes inner connectivity between any point of need and point (s) of expertise. Using synthetic intelligence, the system will capture information already filtered, refined, and interacted with. The bridge 'learns' over time: repeated uses within similar categories will induce the system to begin anticipating (with user approval) requirements of tried and true interaction protocols around specific medical events, etc. Insures operationally appropriate connectivity between expert and need (e.g., making sure EMG output is being rendered via proper EMG rendering systems as opposed to video). Point of need has a characteristic of request (e.g., cardio vs. neuro) which incurs an appropriate response from an expert system. The bridge acts as both a knowledge broker and an educational service. In another sense, the bridge may become able to reach out and create new [kinds of] articulation between relational components, both in real time usage and in data base maintenance. Eventually, it may even be possible and desirable for the bridge to automatically generate a kind of journal of its monthly work outlining new uses and what it learned etc.

3.2. Collaboratory Elements

The next phase development of a functional Med Wide Web is the implementation of Web-based collaborative techniques in the Distributed Medical Intelligence paradigm. These Distributive Medical Collaboratory systems provide necessary structure to the communications which will be utilizing the MWW infrastructure. Web-based collaboratories are the latest emergent functionality to be added to the list of "Web technologies." We are currently rapid prototyping an operational system which will be deployed in the ECU telemedicine testbed. Foundational to this work are some very powerful new web-based software tools and core functionalities being developed by a gifted team at NPAC under the title TANGO, directed by Marek Podgorny (Podgorny et al., 1997). Below we outline the basic structure and function of the collaboratory model and demonstrate its utility with a simple real world example of medical care at a distance with multiple participants and many other functional innovations to the old model of telemedicine.

3.2.1 TANGO

A Web-based medical collaboratory is one wherein a specific group of expert medical personnel, their support staffs, clinical/hospital resources and all related on-line databases comprise the collaborative infrastructure. The collaborative, then, is a networked aggregate of expert knowledge and skill resources deployable as an entity, as needed. "TANGO is an integration platform which enables building useful and deployable Web-based collaborative environments. The system provides the means of fast integration of Web and non Web applications into one multi-user collaborative environment" (Walczak et al., 1996).

Terms

• Application: specific program launched by particular participant (host) which runs on selected-or all-local systems in a collaboratory
• Instance(s): locally active application
• Session: "A session is a group of application instances working together in collaborative mode. All applications belonging to the same session exchange
• information and share behavior" (Walczak et al., 1996)
• Master: "A master is the distinguished user in the current session and has special privileges of controlling the application behavior and/or controlling access of other users to this session" (Walczak et al., 1996).
• Slave: Session Participant
• Browser: Web reader which makes all collaboratory elements functionally intelligible to one another and to its participants; an interface protocol.
• Logging: Data base acquisition of all session activity; machine resident intelligence at all levels exploits this data to ongoingly modify its interfaces, content and prioritization of function.
• Central Server: Analogous to the Bridge, this is the computational brain of the collaboratory. It is the "main communication element. All local demons [systems] communicate with the central server. This server maintains system data (e.g., participants, applications, sessions, etc.). Its main tasks are routing messages among applications participating in each session and recording all events in a data base" (Walczak et al., 1996).

According to the Advanced Telemedicine Technology Roadmap, the first objective is to "Develop a telecommunications infrastructure that is comprehensive, reliable, ubiquitous, and compatible across applications" (Sanders et al., 1996). Once in place, a networked collaboratory of geographically remote participants may look like the following example.

4. Hypothetical Scenario

As an example of the MWW collaboratory in action we offer the following scenario:

1. A remote site encounters a pregnant teenager at risk.

2. Care portal application is set up around the patient (e.g., monitoring devices, diagnostic equipment, etc.) by remote personnel. Via phone dial up, they launch a local TANGO applet. A collaboratory session ensues. The remote site is initial session master in interaction with the bridge.

3. The bridge provides remote access to case-compliant forms in which all relevant patient information is entered.

4. Upon completion of forms, remote [master] site commits data insertion and so re-engages the bridge.

5. The bridge then initiates optimal responses:

• provides open access to patient education content for remote site.
• simultaneously contacts suitable and available participating physician (s).
• opens access to all necessary immediate and archived patient record information for physician (s).

6. Inside a docking station, the physician requests connection with remote site by launching another TANGO application. The physician then becomes master of the session between herself and remote site, which now becomes the "slave."

As a user of MWW what you see is a function of the kind of user you are (what you know), what you are doing, where you are and your need to know.

This simple example hopefully demonstrates the collaboratory's capacity for spontaneous response, emergent functionality, ubiquitous access to relevant resources, and communicative nature supported by the Med Wide Web. In crisis, maintenance or other episodes of medical need this cyber-spatial infrastructure is in place and facilitates necessary expert functioning. By integrating Web-based telemedicine technologies with the computational and system application requirements of TANGO, a powerful and realistic system of Distributed Medical Intelligence emerges: "Distributed Medical Intelligence promotes the development of an integrative medical communication system which addresses the process of providing expert medical knowledge to the point of need" (Warner et al., 1995). As a spontaneous and emergent system it is a network gestalt dynamically changing its shapes with the changing needs of deployment.

5. Final Food For Thoughts

MED WIDE WEB initiates global accessibility and compatibility of intelligent medical communication matrices. MWW will provide world class knowledge access from any point of need, or node on the MWW. Computer supported cooperative work (CSCW) is the future for many highly time critical, information flux based services. It is hoped that the ideas presented here will influence in a positive and empowering way how we (telemedical professionals) conceptualize and integrate Web-based communication technology for the purpose of providing better access to health care, information and knowledge resources. Success with our goals is measured directly in terms of the number of patients healed and given information about their health and medicine in general. The latter is to prevent problems and dispose a patient to being participatory in care given by their physicians.

We leave you with a sense of the relative urgency to act now in ways which will ensure the development of Intelligent Medical Communication systems designed for long term, open-ended and modularizable integration in web-based health care provision. This model is but a small step along a journey which seeks to elucidate possible pathways for technical, administrative, expert and socio-cultural policy so as to brilliantly, powerfully, and globally implement an era of medicine only dreamed about?

Cybercology: "the formal study of cybercultural dynamics as it relates to the techno-social integration of the information society" (Warner, 1997).