Enabling Rehab Researchers: Better, More Timely Access to Information

Interestingly, telecommunications tools can provide rehabilitation researchers with better access to information on subjects. As has been emphasized by NIH panels, there is a need for larger studies that target measures of impairment, disability, functional performance, etc.  The tendency is for researchers to bring a small or moderate number of subjects to a research laboratory, typically about 5-20, due to the challenges of recruiting subjects.  Each group tends to use its own equipment and data reduction protocols, customized for the study.  One of the classic criticisms of such approaches is that they may lack ecological validity, i.e. the insights from the laboratory may not apply to subjects' everyday lives.  Also, rehabilitative processes occur as a function of time, and often researcher use a limited number of samples (e.g., only before and after an intervention protocol).

An alternative is to provide researchers with access to subjects where they live, and to collect data more often.    This could be either as an alternative to a conventional protocol or as a supplement to a laboratory session (e.g., to collect data between laboratory visits).  While for an isolated research team this may seem like more work, if the appropriate infrastructure of validated teleassessment tools was in place, it might actually be easier.  It would also allow new research questions to be asked that are currently not feasible.  These validated tools could be made available through a national repository, as recommended by a recent Workshop on Innovations in Neurorehabilitation that was held at Marquette in November 2001.  New technological tools such as interactive Web sites or telerehabilitation should make it possible to collect data for a large number of subjects (perhaps targeting fewer measures), as well as access to a more natural setting (ecological validity) and to a temporal continuum of data.  It would also help address one of the main challenges in fields such as neurorehabilitation, where the variability between subjects often makes is hard to statistically document training effects.   Larger repositories would also allow emerging tools such as fuzzy logic and data mining to be applied to rehabilitation. 

Another relatively untapped area that will be a focus within this class is unobtrusive telemonitoring.  Such telemonitoring has seen some application for classic physiologic measures such as heart rate, use of accelerometers as activity monitors, use of video and/or audio to monitor persons with neurocognitive impairment, and use of environmental sensors within simulated rooms such as a kitchen.  Notice that some of these employ wearable sensors, while others use sensors embedded within the environment.  But with emerging mobile telecommunications technologies (we will develop these later in the course), the potential is enormous.    

There are, however, several potential barriers.  One relates to limited information content (e.g., an embedded accelerometer gives only a small full indication of degree of movement, and also responds to movement in a vehicle).  Even more important is the challenge of data reduction and interpretation, which the RERC has found challenging, for instance making sense of sensors embedded within a simulated kitchen.  This was anticipated by the Workshop on Homecare Technologies for the 21^st Century, which scored the following at the top recommendation from over 60 produced by 8 expert working groups:

"Fund research for intelligent processing of large amounts of data, including: 1) knowledge assimilation techniques required to optimize the effectiveness of care decisions, 2) information reduction tools for avoiding information overload; and 3) data mining tools for acquiring relevant data from distributed repositories ."

There is a need for greater focus on the process and form for providing information to clinical decision-makers.   This includes approaches for more aggressively reducing and synthesizing data, and better interfaces for presentation.  

Science of Optimizing the Effectiveness of Rehab Interventions

In addition to providing researchers with access to data, there is another key target for scientific inquiry: study of how to optimize the delivery of services. As seen in the figure below, this includes studies of assessments (what to measure, when to measure, etc), therapy (e.g., type, timing, intensity, duration), human-technology interfaces, and issues of compliance/adherence and motivation.

Science of Optimizing Rehab Intervention Strategies.

The use of biosignals to study human movement and performance has a long history, and by 1950 the Biomechanics Laboratory at the University of California had developed quantitative 3-D motion analysis, impressive EMG assessment capabilities, and a solid fundamental understanding of muscle mechanics.  This is noted because it is fair to say that these pioneering research breakthroughs, which predate the development of the allied health field as we know it, have with few exceptions not had a significant impact on clinical assessment and therapeutic intervention, despite an applied health profession that grew up after these developments.  We can collect massive amounts of sensor-based information, but are not sure how to make it clinically useful, in part because human movements are truly complex.  Also, many observations by a highly trained expert practitioner will never be able to be captured by a sensor.  This is why the norm of clinical rehabilitative practice still involves observer ratings and clinical scales for motor tasks, with relatively few objective (sensor-based) measures going into a patient's chart. 

The ideal 21^st-century clinical research tele-assessment tool would be a hybrid that includes both sensor-based information and rater scales, and could be conveniently collected by the researcher since their study could build on an infrastructure that facilitated telemonitoring, such as a national repository of tools.  Hybrid telerehabilitation approaches are needed to help move the rehabilitation field toward the use of more objective measures that meet reliability and validity criteria. We will talk more about possible approaches as this course progresses.

Therapeutic Tele-Interventions

The ultimate challenge in telerehabilitation is clearly to achieve location-independent, integrated therapeutic intervention at a distance, or teletherapy.  Similarly, for researchers studying enabling-disabling rehabilitative bioprocesses, the dream is to be able to apply therapeutic interventions.  Such interventions could range from interactive virtual reality games to monitoring and adjusting home exercise equipment to remotely administering and monitoring pharmaceuticals.   Perhaps more pragmatically, telesupported home therapy could augment lab-based sessions. 

Finally, another top recommendation from the Neurorehabilitation Workshop was that:

"Participants felt that there is a need for low-cost rehabilitation tools that can extend consumer access to supervised therapy into the home. ."

While this was crafted for "consumers" such as the patient or practitioner, it also applies to the rehabilitation researcher. By enabling such access, telerehabilitation tools could provide a variety of new research opportunities for the field of rehabilitative bioengineering.

| tele-intro | tele-history | tele-needs | tele-models | tele-techn | teleconf-standards | tele-science |