The convergence of communication and computational technologies has created informatic systems which may be applied to social and cultural problems. When used appropriately, informatic systems have demonstrated the capacity to enhance the quality of life and to facilitate wellness of being through their applications in areas such as health care and education. Interaction between informatic systems and systems which they influence can be qualitatively and quantitatively modeled by generalized theories of adaptive complex dynamical systems. Interventional Informatics is the proactive utilization of informatic systems as complex adaptive systems to alter the course or outcome of a particular behavior.

Interventional Informatics

We live in the information age. Our lives are continually influenced by new applications of information technology. Informatic systems impact society at all levels, from personal and interpersonal dynamics, to global systems of immense complexity. Informatic systems have become so complex that their implementation causes certain unexpected and often undesirable behaviors to emerge in the individuals or groups involved. This often evokes a shortsighted linear reactive response in an attempt to gain some predictability, though this usually results in a worse state than before (take the public education and health care systems, for example). There appears to be a fundamental lack of understanding of these systems when they go beyond a certain level of complexity. Is there a way to resolve this?

A Suggested Solution

As part of our eternal journey to make the world a better place for expression at all possible levels, we recognize the need for a basic theoretical understanding of the dynamic relationship between information technology and society. If we understand how information technology is capable of affecting our lives, we may respond proactively to avoid some of these affects and to cause others. More commonly we find, however, that we have a need to understand information technology in order to respond reactively when it behaves unpredictably (understanding rarely means 100% predictability). In a reactive state, we look for critical points, state transitions, where familiar patterns may manifest in some system at some level and which hopefully will provide a meaningful link with an existing model. A major goal for complex information systems such as education and health care should be to reduce the need for a costly reactive response by taking proactive steps with information technology. With this in mind, we have coined the term Interventional Informatics (II) to describe the pre-emptive, proactive, or preventive use of relatively small amounts of information and information technologies at critically sensitive points on a system's information state trajectory.

Theoretical Foundation

Most readers are familiar with the term "intervention" as it applies in a social setting (addiction therapy), or in a political setting (first world intervenes at a critical state in a third world country), or even a pharmacological setting (drug therapy). We may even call a vaccine an immunological intervention. Even though these all involve some form of information and they consist of acting either reactively or proactively at a potentially critical point in a system's evolution to hinder or alter an imagined outcome, they do not necessarily depend on informatics systems to operate normally. Neither, for example, does the interventional act of informing someone that they may reduce the risk of AIDS by using condoms depend on informatics. Rather, interventional informatics involves putting an information-based mechanism in place which understands and uses the tremendous potential of small amounts of information, when applied at critical junctures over a range of scales, to alter outcomes in a positive way. How do we transition from a reactive to a proactive state? We are now at the point where we can benefit from applying principles of complex adaptive dynamical systems to large systems like the global communications network (Internet), although probably not quite yet with the same quantitative precision of neurophysiological or biochemical research. However, we are drawn by a strong qualitative similarity in the dynamics of seemingly different systems. To understand why this is so and what use this might serve, it is essential to recognize that at the very least, the behavior of many complex spatially-distributed dynamical systems appear to provide a metaphorical basis for some powerful interdisciplinary languaging tools. These tools may help us identify system behaviors familiar to us in other more well understood systems, at which point more formal methods of analysis may be applied.

Social Systems and Their Respective Levels of Complexity

It is useful to distinguish levels of social complexity at which information technology's importance manifests. They include: the individual, interpersonal (intimate/non-intimate), group, community, district, city, county, state, region, country, continent, hemisphere, and planet. In academic medicine, it is also useful to consider a common non-intimate interpersonal system which manifests in different contexts but exhibits similar behaviors. It consists of one person who is being assessed and the other doing the assessment. These are: the physician /patient, the physician /student, and the student/patient. To a less obvious extent, there are also the License Commissioner/physician or practice, and the Medical school Accreditor/Dean et al. In each of these systems, interaction is bi-directional and, contrary to popular belief, when functioning properly the boundaries between the assessor and assessee are blurred. Each of the two system components should be providing informative feedback to the other about their performance, and adapting accordingly. These are reasonably simple examples of complex adaptive systems whose behavior differs in detail, but are quite similar in how information technology may be beneficially applied. They each involve a process of learning and assessment through informative feedback. In order to maintain proper function, it is critical to determine how information technologies are best applied to these systems.

The Role of Information Technologies in Assessing Learning and Performance

In general, improved retention, recall, and understanding may result from either of two basic processes. The first involves improved fidelity of the information being processed, i.e. a more accurate rendition of an existing representation. Examples would be modeling and simulation of a system, or in academic medicine, the case-based curriculum. The other learning process involves creation of a rendition richer in information because of changes in the nature of meaningful associations made with an existing rendition, perhaps through an increase in clarity, strength, or number of associations. This process includes principles of dual-coding, perceptualization (see below), and active learning. Interactive information technologies (i.e. any technology which responds in a nonarbitrary way to input from a user, and which in turn evokes a response from that user), and especially virtual reality technology, may potentially merge these two learning processes and serve to emphasize the important role of creativity in both information-rich (rule-based scientific) and entropy-rich (artistic narrative) behavior. In this regard, it is particularly important for our children to become adequately informed and capable of expressing themselves and informing others at all levels. Given the opportunity to develop their own expressive abilities and desires unhindered, we are certain they would turn both the education and health care systems around, and many others as well.

Case Studies in Interventional Informatics: Work In Progress

Recently, we have come to know an eight year-old girl named Ashley and a nine year-old boy named Trent. Ashley, who has been a C1 spinal (high quad) nearly her whole life, is an amazing young girl. She lives with her grandparents, and it was they who contacted us after seeing a television broadcast about our work with virtual reality and the disabled. Trent has only recently been physically challenged. Since we first met, they have learned to play video games, interact in meaningful ways with 3D objects in a virtual world, and drive a car, all with just their faces! By integrating off-the-shelf computer and data acquisition technologies with custom interactive software and hardware, we have created a system in which an individual may control their environment through their body's bioelectric signals , such as their muscles, eyes, and brain. Figure 1 represents a system in which a user may use their ability to control these signals in order to remotely control a robotic device, in this case the speed and direction of a small battery powered car equipped with stereographic cameras, microphones, speakers, and lights. The car receives visual and audio information from its environment, transmits it to the user, and receives controlling signals and audio from the user in response. Where is the informational intervention in the process of Ashley and Trent becoming more empowered to learn? Figure 1 represents an information trajectory which we have found to be useful in identifying system components and key points of intervention. The trajectory delineates the informational basis of a cognitive cybernetic loop. It is divided into a path representing the direction in which the user's abilities are manifested, and a path representing the motivational feedback which makes the interaction a reality.

Ability

In Figure 1, the path representing ability contains one of the most important points of intervention, the bioelectric controlling software (BEC). This kind of informational intervention at the device level is more basic and inconspicuous than other levels, but no less critical. This software is used to calibrate the muscle signals and to define complex sets of gestures for various navigational tasks. These gestures suggest the term cyberlinguistics, and they may be looked upon as a sort of 'cyberpidgin'; a collection of functional but grammarless phrases useful in commonly occurring interactions. It is possible, even likely, that these pidgins may develop into Creoles and ultimately full-blown languages as the need evolves. This suggests the potential for information technology as a foundation for some remarkably unique languaging tools. The task is to develop the cyberlinguistic navigational heuristics to use them. The successful operation of the system by Ashley or Trent hinges on the adaptability of this software. It is not yet auto-adaptive (it must be configured semi-manually), which would be yet another way of intervening at a critical point to enhance a user's expressive capacity.

Motivation

In rehabilitation, as in other types of learning, motivational feedback can be critical to rapid therapeutic progress. In the motivational component of this system, intervention is critical in the choice of rendering modes. By selecting binaural (3D spatial perception) and stereographic display modes, we have committed to mediums significantly more faithful to the source than text or 2D sound and video, and which provide a foundation for increased perceptual associativity. Ashley and Trent have become mobilized spatially-distributed vehicles of sight and sound. Through their learning and adaptive skills with their face muscles, they are now spatiotemporally extended well beyond where they were only a very short time ago. And of course, with the proper interactive informatics toolkits, these navigational skills may be used in a much more abstract way than simply driving a car. We have not yet given Ashley or Trent the opportunity to meet together in a world of their own making, but someday soon we hope to. Thanks to Trent and Ashley and a few others like them, in just a short time we have learned a great deal about complex adaptive systems; just look at their faces while they're playing.

Modeling Complex Adaptive Behaviors

There are several properties of complex dynamical systems which may be useful in modeling the effect of information on society. Two important properties are nonlinear response and control parameters. The property of nonlinearity implies that the magnitude of a system's response to some input is not directly proportional to the perceived value of that input, in fact it is usually far from linear. Additionally, the state of the population prior to the input of information strongly influences the response of the population to that input. There exists certain parameters in the population which strongly bias the response potential for a given input. These control parameters may be modified either internally or externally, thus providing a choice of mechanisms for altering a population's response potential.

Useful Similarities in Apparently Different Systems

Let's consider two systems as examples, (1) the global information networks such as the Internet, and (2) bioelectric-controlled telepresence such as Ashley using electromyographic (EMG) signals from her face in order to perform telerobotics. Control parameters in the Internet might be the number of on-line information servers globally, or the number of users with hypermedia-based network browsing programs like Mosaic or Netscape, and a measure of complexity (an order parameter) might be the average rate of information flow. Analogously, the number of sensors capable of being attached to Ashley's face, or the number of gestures available simultaneously for her to manipulate might serve as control parameters in that system. As we vary these control parameters, the complexity of the systems evolves, and we expect to observe rapid, distinct, and highly nonlinear transitions in their behavior. Evidence supporting this has been observed in the rate of increase in traffic (information flow) on the Internet, probably as a function in part of the control parameters mentioned above. Traffic jumped in such a way in the beginning of 1993 when the first multiple OS World Wide Web browsing program (NCSA's Mosaic) began distribution1. We suggest that a corresponding situation exists in bioelectric telepresence, such that increasing the number of degrees of freedom (sensors) available to Ashley or the combinatorial capacity of the gesture definition software should lead to critical transitions in information flow. We have not tested this empirically, although such behaviors have been observed in learning procedural tasks similar to those in this system2. To do this with Ashley or Trent would require an extended study involving in part some intentionally controlled limitation of their capacity to interact with their world. This kind of study may hopefully be an easier one to consider in the future, but at present we feel it is not worth the aggravation and frustration such constraints place on children who are already constrained by their disability.

The Future

Where else might interventional informatics be applied? We used information technology proactively by anticipating the general need for Ashley and Trent to communicate with their families and friends, but we were surprised at the creative behavior of siblings and friends and what it implied: that not only were Trent and Ashley more empowered to interact with their world, the car provided part of Trent's world, his siblings and friends, with a "vehicle" to interact with him, too! They were jumping in front of it, picking it up and flying it around. This sort of unanticipated behavior may emerge in other ways at other levels, and it should be accounted for in an accurate model of the system. We are actively exploring the areas of somatosensory and vestibular experiences, although we are still in the early stages. Many of those who are confined to a wheelchair lack much of the usual variety of these perceptual experiences. We are exploring how immersion in low frequency high amplitude sound, using speakers external and internal to the equivalent of a high tech Lazy Boy, can cause a meaningful somatosensory-induced deep visceral experience. We are also attempting to combine this with a vestibular component using motion platforms or suspension systems. These efforts represent our emphasis on perceptualization, which we define as the integrative experience of consciousness as the primary rendering state. The term originated in response to the severe visual bias so common in information technologies, and it refers to the integration of multiple sensory processors and their basic spatiotemporal pansensory features such as frequency, intensity, and complexity. On a more global scale, the telecommunications media have provided us with an initial link to some of those in society who could dramatically benefit from these technologies, but this is not a reliably predictable medium for a formal informatics mechanism. Our ability to manipulate the telecommunications media in it's current state is severely limited as a result. It is critical that a more formal informatic infrastructure be put in place which virtually eliminates any reliance on luck to reach patients in need. The information superhighway offers promise towards this goal. It certainly allows us to test some of the most important ideas about complex adaptive systems. Interaction occurs on the Internet at many levels, with varying complexity, but with a common purpose, to proactively share information. The concept of "pulsing the net" at a critical state with information evokes images of echolocation, cybersonar.

Conclusion: Dynamical Similarity at Varying Social Scales

Finally, at a more basic and personal level, if we recognize that each one of us has emergent adaptive qualities in common with our environment, we may more readily recognize critical points in information flow between components of a system, and across systems of varying levels of complexity. Herein lies the essence of proactive informative behavior that we observe with the bioelectric controlling software, and in the global information infrastructure, and in many other systems dependent at least in part on information technology. Our relentless need to communicate, combined with an understanding of these technologies, as well as a recognition of the abilities and critical state of relevant social systems, should result in a dramatic increase in expressive capacity for society at all levels, all towards making the world a better place.