Distributed Diagnosis and Home Healthcare (D2H2)
Bioengineering--the discipline at the interface between medicine and engineering--is poised to begin a reinvention medicine. This new medicine will be based on the explosion of new knowledge of the molecular basis of life processes, and new tools for the manipulation of the genes that orchestrate the activity of those proteins. It will be implemented with instruments we are now developing that will change where medicine is practiced. Ironically, the new medicine enabled by this technology may, in an important sense, be less be less of a giant leap forward than a small and welcome step backward.
From the little black bag to the urban hospital center
When many of us were young, the doctor was someone with a black leather bag who came to our house when we were ill. He consulted with us and with our family members, examined us, prescribed and even gave medication, all in the home. Families of several generations participated in all the stages of life, from working with midwives at the birth of children in the home, to easing the death of the elderly, at home with their children and grandchildren. However, we have always wanted to give ourselves and our loved ones the best and longest life that money could buy. Already by mid-century we had nearly abandoned the model in which the general practitioner ministered to us from birth to grave in our homes or in a small office near to where we lived. We had made great strides in inventing new technology for the betterment of human health. This technology, whether it was the x-ray generator, the iron lung, the blood diagnostics laboratory, the CAT scanner, the MR imager, the arthroscopic surgical suite, or the modern ICU, was wonderful in the power it gave the medical practitioner, but it was, complex, fabulously expensive, and in need of constant maintenance by many highly trained technicians. Nowhere but in a large centralized hospital could you afford to have all these tools. To have good medical care, then, in the latter half of the 20th century, you had to be in a hospital. If you wanted your children to have the best chance of being born healthy, you delivered in a hospital. If you became sick, you went to a hospital and stayed until you became better or died. If you were old, and unable to care for yourself, you were sent to a nursing home, and when you died, you died in the care of those who were technically the best trained to oversee the process, but away from family and home. By the end of the 20th century, we have, in the name of improving medicine, remanded our care into the hands of specialists, hidden from view behind the walls of institutions that can afford our best technology. The Egyptians built their pyramids to preserve the dead. We have built ours to prolong the lives of the living.
One can argue that the social trends like fragmentation of families, mobility, and changes in the workplace have played a stronger role the changes in our culture than has the change in medical care. But for many people choices of lifestyle and location that disrupt families have been available because of the very quality of medical care available for the young and the elderly. The practice of medicine is inseparable from the culture in which it exists.
The analogy of computation
If we accept the possibility that our relentless pursuit of the centralization of medical care has not uniformly improved our lives, is there an analogy that suggest a way out of our present situation?
Computers were originally people, eventually assisted by adding machines. After WWII there was explosive technological advance in large fast computers, leading to more and more centralization in computation, culminating in the era in which all computation was centralized into a few facilities. Computation came under the exclusive control of government and large corporations. If you didn't bring your punch cards to the computer, your problem wasn't solved. The second wave of computer technology, the microcomputer revolution, has dropped of cost of the hardware to the point that we can all afford machines more powerful than the best machines from the time of the height of centralization. Thanks to the computer, the software, and the wired and wireless data communications network, we have now created a "distributed computing network", the apotheosis of which is the World Wide Web. We are all, now, computers in the pre-WWII sense, but ones with powers unimaginable at the beginning of the 20th century. We can formulate problems we would never have considered, and receive the answers to our questions in tiny fractions of the time that would have been required to ship our punch cards to the central computer. This decentralization of computing power has made changes in our society far beyond affecting how we do mathematics. We are now beginning to live in a world with virtual corporations, telecommuting, and a society in which we can live and working in technical jobs while living in homes distributed outside of major metropolitan areas. The distribution of the computational power has changed how we live.
The future of medical care
We see the decentralization of medicine as not just bringing back the little black bag and the house call, but empowering us as individuals to take over technical responsibility for some aspects of our own medical care and that of our families. If we accept the proposition that such decentralization is a good thing, can it be accomplished? Can we take medicine and decentralize it in the same way that computation was decentralized? What technology would be required to accomplish this? What infrastructure would have to be put in place to support that technology?
Distributed Diagnosis and Home Healthcare (D2H2)
We foresee that in the next decade or two, new technology will allow us to effect the same sort of change in medicine as has occurred in computing. With one set of technological changes, we can 1) reduce some key inefficiencies in the current practice of medicine, 2) allow improved care by allowing continuous measurements of biomedical information outside the hospital, and 3) greatly improve the quality of the lives of those of us in the developed and developing worlds.
Several types of expensive centralized biomedical instrumentation can be miniaturized and made inexpensively enough that the diagnostics they provide can be distributed to sites outside of where they are currently used. The ones where we see immediate potential are in diagnostic imaging and in biochemical diagnostics.
Diagnostics available in point-of-care scenarios
We foresee a two-stage process of decentralization. The first is ongoing, and involves putting more portable and handheld instrumentation into point-of-care environments. For example, there is now a handheld instrument manufactured by I-Stat of Princeton, NJ, that allows simple blood chemical measurements to be made in the ER and by the patient bedside. The use of this instrument has shortened the delay in obtaining blood chemistry measurements from longer than 20 minutes in even the best hospitals to less than 1 minute. In some emergent conditions, this type of timesaving is important enough that whole hospitals have been outfitted with such devices. Some doctor's offices have also adopted this instrument, rather than waiting days to get the same information by sending it out to a laboratory across town, or, in some cases, across the state. Nursing homes with trained staff are beginning to think about adopting some similar technology.
Diagnostics available in the home
The next phase, and the more revolutionary one, is putting the same sorts of capabilities into the home. We already have home pregnancy testing, and portable blood glucose testing. With an aging population, new populations of people on long-term multi-drug regimens, and increasing numbers of drug-resistant infections, there is a clear niche for development of a home instrument that could provide immediate specialized medical information to the user, either directly or through a central health care provided acting on the basis of the information generated in the home.
One simple goal would be to see to it that a person would not have to drive 50 miles to the emergency room at 3 a.m. on icy roads to find out if whether he is having a heart attack or indigestion. There are blood markers that appear early in the course of a myocardial infarction that, combined with discussion with a health care provider, could make the decision early and with little risk to the patient. To implement this we have to be able to place small inexpensive fluid-analysis instruments in homes, ambulances, and extended care facilities connected via a local PC to the Internet for rapid transmission of chemical data to centralized diagnostic facilities. We envision a self contained box no larger than an external removable disk drive (and with no more complex maintenance) attached to the home PC. The potential patient or a local caregiver, friend or relative, would begin user-initiated routine and critical testing of multiple panels of blood, urine, and sputum chemicals based on permanent and disposable elements available over the counter. The results of the test need not be displayed to the patient&--;two-way audio and video communication will be established between the test site and expert care providers who can interpret the chemical data and recommend action on the part of the patient and, if necessary, of an ambulance crew. The anxious patent is told by audio and/or video link within minutes to unlock the front door and wait for the ambulance, to take two antacids and see the physician in the morning, or to increase his valium dosage and go to bed. Cost savings will accrue through earlier diagnosis and, thereby, more effective treatment of chronic and emergent medical conditions, e.g., biochemical imbalances, toxic drug interactions, cancer, heart attacks and strokes.
The societal impact in the US
These developments could be described as a form of telemedicine&--;a concept that has been discussed extensively in the medical community. We foresee that such instrumentation, if it could be provided at a cost comparable to home PC could become as indispensable to the middle class American family as the PC has become. The permanent instrument would work in conjunction with condition-specific disposables that included reagents and waste disposal components. These would both be distributed at the neighborhood pharmacy. The instrument would carry (or download on demand) enough on-board capability to self-calibrate and quantify the technical information acquired, and provide immediate unambiguous recommendations for action to the user, if required, and ship the same data to a remote site for interpretation, filing, and recommendations by a combination of expert systems and care providers on call.
The potential for a worldwide medical network
Ultimately, if such a D2H2 system and the associated instrumentation could be provided at a low enough cost, it should allow distribution of medical care to parts of the globe where very little medical care is now available. In those places the institution of a global system would bring a real revolution in the quality of medical care.
What is available now?
The good news is that much of the technical work for developing a system such as this is either in place or under development for other purposes. The establishment of global high bandwidth networks, with local links capable of handling digital information is well underway. Most of the medical information needed is known, and the chemical and physical methods are well established. Furthermore, existing and planned global wireless and wired data communications networks should allow all the bandwidth required to transfer the combination of chemical, voice and image data required. The PCs exist as a platform on which to base the instrumentation, and convenient methods for distributing software updates via the Internet are in place. Furthermore, the US has ample numbers of trained physicians capable of assisting in establishing the network required, and capable of providing the expert personal input when such personal involvement in cases proves necessary.
What needs to be developed
There are pieces missing, however. One is a network for 1) accepting and processing the medical diagnostic information from wherever it was generated, 2) presenting it to a combination of expert systems and physicians, 3) directing instructions action items to the patients outside the hospital or clinic, and 4) ensuring continued safe and reliable operation of the data collecting instruments at the point of measurement. These four tasks will be required whether the goal is a near-term system for operating such instrumentation in doctors' offices in the US or a global home-based medical care system. One can argue that these sorts of functions are best handled by those companies already in possession of the lines of communication. This will require the involvement of major corporations deep pockets and long-term vision.
Instrumentation available at the point of use&--;the role of microtechnology
The other missing piece is the instrumentation required to create the physical and chemical data in the first place. The charge is currently being led by DARPA DSO and ETO. It has lately been spending millions of dollars to fund of universities and industry for development of Microelectromechanical systems (MEMS) and MEMS-derived portable diagnostic instrumentation. These instruments have been designed both for detection of chemical and biological warfare (CBW) agents and for combat casualty care. These programs have, over the last 5 years, led to an explosive growth in interest in what can be done to convert the microfabrication technologies designed for the microelectronics industry into tools for measuring chemical and biological samples. Whole new fields such as microfluidics, composite CAD, and micro-total analytical systems have sprouted. The goal is the development of portable rugged versatile instrumentation for chemical and physical monitoring of environmental and biological samples. They are designed with the DOD emphasis on making remote data available to the command and control centers, so creation of two-way data transfer systems are integral to the development of the data generation devices.
The DOD focus of the DARPA-funded research has not prevented the contractors from forging the microanalytical tools into products for the domestic market. The private sector is following with rapid implementation of these techniques for instrumentation for DNA sequencing, combinatorial chemistry and high throughput screening. Many new companies have sprung up around DARPA support for this type of work. A few of these companies have targeted point-of-care diagnostics as a focus for their near-term product development. The instruments under development are generally designed as stand-alone devices, but they clearly could be used as the basis of the D2H2 network.
The opportunity in Bioengineering at UW
We believe that there exist now a unique opportunity at UW, and particularly in the Department of Bioengineering, to become the center for research and development of Distributed Diagnosis and Home Healthcare. There is already a strong base of existing research in Bioengineering in the technologies that will enable D2H2. The work is occurring both in Bioengineering and in affiliated department through collaborative projects. Briefly, these are as follows:
Microfluidic Chemical Analytical Systems
Biomedical diagnostic technology must be inexpensive, chemically versatile, and relatively accurate to be successful as a commercial product. The size of samples of blood will have to be on the order of a few drops. This immediately brings us into the realm of microfluidics. Practical implementations of microfluidics require dealing with samples far more complex than those regularly introduced into instrumentation with such narrow channels. Microfluidic systems are very vulnerable to problems inherent in unrefined samples to be encountered in the scenarios for which these instruments are being proposed. For example, many of the best sensing technologies in the analytical chemist's arsenal are not suited to complex mixtures of analytes with overlapping interfering signals. This is true for both optical and electrochemical detection methods. Surface fouling is a many-faceted problem in small channels, in that it can lead to converting every fluid transport channel into a chromatography column, or, in the worst cases, can lead to loss of all of the sample to an irreversibly adsorbed layer upstream of the detector. Electrochemical sensors have always been vulnerable to fouling, but in microfluidic systems, the electrodes will usually not be removable for refurbishment by the user. Some types of ancillary techniques, such as electroösmotic pumping, dependent as they are on the nature of the chemistry of the channel walls, are particularly susceptible to disruption by sample-to-sample variability.
Portable ultrasound imaging
The University of Washington pioneered the research and development of a portable ultrasound device for battlefield trauma with ATL via funding from DARPA under the Technology Reinvestment Program. This UW initiative contributed to the development of the SonoSite (an ATL spinoff company) and its 7-pound hand-held device, which will be commercialized in 1999. Even though this is not designed for home use, we believe that they will be introduced for home use in the next decade with decreasing costs and improving usability. In the beginning, they will be used for imaging the body and transmitting the resulting images to the clinic via telemedicine. However, they could be eventually used in both diagnosis and therapy.
Image-guided therapy offers the potential to direct therapeutic action precisely to the point in the tissue where it is needed and not to other tissues. When this is possible, a higher dose can be administered with higher probability of a complete cure. Focused delivery requires the integration of three essential components: (1) Accurate and clear imagery to identify the offending tissue, (2) a therapy that can be accurately directed and controlled, and (3) a well-controlled means to guide the therapy to the imaged location. High intensity focused ultrasound can offer exciting new therapeutic methods to ablate offending tissue or stop hemorrhaging vessels and organs. The action is thermal and mechanical. Selective therapeutic action can be enhanced with energy-activated drugs, and particularly those which can be combined with contrast agents and/or directing antibodies. This technology brings a new type of therapeutic tool to bear on injuries that must be treated before the patient can be transported to a hospital.
Development of imaging hardware and software
Researchers at UW have worked with Texas Instruments to design the TMS320C80 Multimedia Video Processor (MVP). The MVP is still the most powerful image processing chip and being used by many companies, including Sony, Tektronix, Siemens and General Electric. Using the MVPs, we also developed a programmable ultrasound image processor for real-time ultrasound processing, which is widely used in Siemens ultrasound machines. It provides a platform for rapid testing of new concepts in ultrasound processing and enables software upgrades for future technologies. Current clinical applications include generating a panoramic view in real time (called SieScape), quantitative measurements, 3D ultrasound, harmonic imaging, adaptive persistence, image segmentation and telemedicine. For example, a reduction of overall execution time of an ultrasound fetal-head detection algorithm from 32 s (on SparcStation 20/71) to 248 ms was achieved. We also used UW-developed MVP-based systems for advanced telemedicine between the University of Washington and Madigan Army Medical Center in Tacoma in 1995 for transferring X-ray and live ultrasound images using a 45-Mbps ATM switched network. We are currently completing the next-generation multimedia processor, which will be commercially available in 1999. Our simulation results indicate that this processor is about 4 and 20 times more powerful than the MVP and Intel's Pentium II, respectively.
Other UW units supporting the development of D2H2
Two other UW Bioengineering enterprises, the NSF ERC entitled UWEB (University of Washington Engineered Biomaterials), and the NIH Center entitled NESAC/Bio, are the foci of decades of multi-million dollar support from the federal government to develop novel surfaces and equipment to determine what is on those surfaces. This concentration of expertise and funding gives UW an absolutely unique ability to solve one of the tougher problems in developing such biomedical instrumentation&--;development of chemically and physically appropriate surface coatings.
We also have on campus two entities aimed at promoting the interaction between on-campus research and the industrial sector. The Washington Technology Center (WTC) is a state-funded entity that has, among its many activities, created and funded a large and user-friendly Si microfabrication facility at the UW campus that is available to both UW researchers are companies on a fee-per-use basis. They also have begun to actively support MEMS-related research through targeted funding. The newly created Center for Applied Microtechnology (CAM) is a clearing house for information about MEMS and microelectronics related information meant to serve both the UW and the industrial community, particularly in the Puget Sound area. Both CAM and WTC work closely with the Department of Bioengineering, and provide the fabrication infrastructure for R&D for UW and industrial partners.
To develop the new molecular chemistry and physics that will be the basis of the next generation of molecular diagnostics, the UW last year funded the Center for Nanotechnology with its University Initiatives Fund. This program is multi-department, but is based in Bioengineering, and along with the Molecular Bioengineering pathway in Bioengineering, is responsible for training a new cadre of engineers who will trained in modern molecular science.
Return to Microfluidics Home Page
Return to Yager Group Home Page