Integration of MEMS and medicine will have pervasive impact

Sounds like a fascinating topic. Where are we right now?

Paul Gourley: What we're seeing here in the last four or five years is the development of a new technology, which is really the marriage of microbiology and microfabrication (borrowed from microelectronics), perhaps more specifically, MEMS (microelectromechanical systems) and nanotechnology. These two fields are starting to impact the areas of microbiology for clinical diagnostics, environmental sensing and monitoring, and will have a pervasive impact on basic research in molecular and cellular biology, all arising through interdisciplinary work in materials engineering, biology, physics, and chemistry. The term "microfabricated chip" conjures images of silicon wafers, but we're discussing a more diverse set of materials that include semiconductors, glasses, silicone elastomers, and polymers. Further, various means for marrying different materials are being investigated. Despite the differences in material technologies, these chips are similar in that small features (grooves, channels, islands) are microlithographically defined to enhance device performance.

Can you give some examples?

Paul: One notable example is the new company Affymetrix, Inc. (Santa Clara, CA), which has used microlithography techniques to make a glass chip that can analyze specific DNA fragments. This so-called Genechip (Figures 1 and 2) can help determine whether a young woman is a cystic fibrosis carrier, whether a patient has AIDs, or it could analyze sperm cells and tell for court purposes whether a young man is or is not the father of a child. Things like that which, when you think about it, have pretty far reaching effects. This new company is now producing these chips, and it's expected that there will be a fairly large market for them in the near future.

Figure 1. (i) Genechip probe array in a cartridge. (ii) Labeled schematic of the cartridge. Two septa (a) serve as ports for fluid introduction and removal; the chip (b) is assembled with the oligonucleotide probes on the inside surface of the hybridization chamber (c). The cartridge includes features for alignment in the instrument (d). Reprinted from American Laboratory, Volume 28, no. 5, p. 39. ©1996 by International Scientific Communications, Inc.

Figure 2. The Genechip scanner was developed to detect a fluorescently labeled target hybridized to the probe cells on the chip. Incident light from an argon ion laser (a) passes through an in-line filter (b) and is then directed from a dichroic mirror (c) that is focused on the scanning head (d) via a routing mirror and confocal optical system. The chip cartridge holder (e) moves relative to the beam, which excites the fluorochrome with a spot size of approximately 8 µm. Emitted light passes back through the optical system, including the dichroic mirror, where it encounters a set of filters (f) that allows only light of the appropriate wavelength to pass through. Remaining light is focused through an achromatic lens (g) and pinhole aperture onto a photomultiplier tube (h) that converts the fluorescent light into electrical current.

Are these chips outside or inside the body?

Paul: These would be outside. This would be a clinical application where DNA would be extracted from blood or other cells and injected into this Genechip.

With fast results?

Paul: Yes.

Mark Gourley: To pick up where Paul left off, the ability to sequence DNA is in the headlines of newspapers and television whenever there's a notorious court case, but the ability to rapidly sequence DNA is not available. Traditionally, when you sequence DNA, you first have to isolate the DNA by putting it in a gel to smear it out so you can count each individual base that makes up the entire DNA.

Do you have to determine the entire DNA structure for court cases?

Mark: No. If you don't have enough DNA to do that, then you amplify it through the well-known polymerase chain reaction. With the ability to sequence DNA on a chip, we can use a very small amount of sample and sequence the DNA more rapidly.

So, this new technology promises to be more economical also?

Paul: That's right. You're leveraging the power of semiconductor microelectronics, built up over the past 20 years or so because of computer technology. The advantages include small feature dimensions for making huge numbers of elements, so one can insert or inject small fluid volumes in parallel. They would have high sensitivity through materials engineering and use small volumes of analyte. There is stronger signal to noise, because background noise is minimized. There's also higher speed; you can get information more quickly.

Because these microsystems are so small, many of them can be fabricated on the same 1-in. glass or semiconductor chip. So, one has a massively parallel operation in which one can simultaneously probe for say DNA fragments, maybe thousands to millions, by using laser scanning or CCD imaging methods. So, one obtains considerable information for low cost. And these chips can be highly manufacturable, marketable, and portable.

Mark: From the clinical side, the big advantage is not having to process or prepare the sample for analysis. In a clinical sense, you can analyze things in real time. Bringing the technology into the clinical arena allows the real advantage of not having to process the sample, which requires a lot of time and expense. Furthermore, without the need to process the sample, we have the ability to analyze it in real time and study cellular function as it occurs before our eyes.

I foresee a lot of this new technology will enable your local lab to do the work rather than sending the specimens to more specialized labs.

Mark: Medicine has changed in that physicians tend to send out samples to a main reference laboratory. I envision, if this type of technology is to the point where it becomes simple and accurate, that this would be a procedure or process that could be done in a clinic. You could prick the finger of a patient, put the blood on a chip, put the chip in the instrument, and obtain your answer. These systems could be widely distributed in clinics or doctor's offices, maybe even in homes.

How far down the road is this?

Paul: There are companies that have started up and are actively developing these products. I'm expecting that we should start to see the impact of these devices on the research community within the next year or two. Of course, clinical application for human diagnostics requires FDA approval and that will require more time. Eventually, I expect to see a very pervasive role of these devices in diagnosing diseases at early stages, detecting genetic abnormalities, and speeding the development of drugs.

Do you have a sample chip?

Paul: Mark and I have invented a new kind of cell analysis chip, based not on silicon, the prominent semiconductor material being used, but instead on the marriage of a compound semiconductor GaAs, which can lase, and glass dielectrics.

What's the advantage of GaAs?

Paul: Its light-emitting ability. Once you have the material on the chip, you somehow have to be able to detect its presence. Usually, that's done by laser-induced fluorescence of a molecular tag or some other optical technique. The chip is a flat surface; by using optics, one can probe many different regions of the chip.

Figure 3. The upper left figure shows cells placed in the laser cavity formed with semiconductor and dielectric mirror surfaces. The semiconductor is photopumped by a separate laser to generate electron-hole pairs that recombine to emit light. The cells act as transparent waveguides to channel the light between mirrors and aid the lasing process. Stable optical modes confined by the cells, illustrated in the lower left diagram, will support lasing. Near-field lasing images are shown on the right side for normal and sickled red blood cells. Also shown are corresponding lasing spectra that span 845 to 860 nm on the horizontal wavelength scale and 3 orders of magnitude on the vertical logarithmic scale. The spectra provide unique signatures and can be used to identify the cell type, size, and shape.

In our technology, called a biological microcavity laser (Fig. 3), we read the individual cell information from a laser beam that emanates from the chip. The cells are injected inside a laser microcavity formed by highly reflective semiconductor and glass surfaces. When the semiconductor is photopumped or electrically activated, the cells act as waveguides to aid the lasing process. Thus, the cells light up intensely against a dark background. Information about the cell size and shape are impressed spatially and spectrally onto the beam. The transparent cells don't need to be stained; the dielectric properties of the cell modify the lasing spectra. This technique also allows you to assess many cells on the chip at a very high speed and exploits the advantages of laser techniques, like interference spectroscopy, ultrafast spectroscopy, or parallel ray detection.

An example of our use of the microcavity laser is given in Figure 4.

Figure 4. Human lymphocyte supporting lasing modes. The upper left image is a laser scanning confocal micrograph of a lymphoctye before sealing into the laser cavity. The three images and spectra below are recorded with the cell in the laser cavity. The bottom spectrum is recorded below the lasing threshold and shows only broad spontaneous emission. The top spectrum shows lasing modes associated with the nucleus. The middle spectrum shows lasing modes associated with the cytoplasm and nucleus. The peaks provide information about the distribution of DNA/protein complexes and composition of the cell.

So, this is a smart chip. Other than your laser, what other components do you have on this chip?

Paul: We have means for injecting and transporting fluids, in addition to the laser analysis. Ideally, people would like to see a chip that has onboard all of the following components: (1) some means of collecting the sample or means for injecting it, (2) a means for concentrating or preparing the sample for measurement, (3) a means for separating the various components of the analyte, (4) the ability to do the analysis or detection, preferably, with a laser-based scheme using optical detectors, and finally (5) some means for processing the data to give you an output.

No one has really pulled off all of this at the present time. People are working on various components of it and various projects are being funded, by private industry and government, for these laboratories on a chip, or sometimes, chem lab on a chip, as they're called. The promise of use in health care and environmental monitoring is driving this.

What kind of dimensions are you talking about for this chip?

Paul: The dimensions might be as large as the wafer itself, a 6 in. diam. That would be a very sophisticated chip. Or it could be as small as perhaps a few hundred microns.

In addition to these various on-chip schemes, there are hybrid schemes where one uses some of the analysis on the chip and then has other, say, larger components that follow or precede the chip. But we feel that the more functions you can achieve on the chip, the simpler and more efficient the process will be. Semiconductors with electrical and optical functions look especially attractive in this regard.

When you mention nanotechnology, I envision these very small devices inside the body measuring properties. Is that what you guys are also envisioning?

Paul: There are several variations. One use of nanotechnology is in the clinical diagnostic sense, using small-scale structures on chips outside the body to do analysis. But, you're right. You can also use nanotechnology in the body. In fact, one of the papers at the SPIE meeting (BIOS '97, San Jose) discusses using magnetic particles, about 200 nm in size, that are injected into the body. These particles would collect at tumor sites to be better imaged with MRI techniques, thus helping pinpoint tumor locations.

It seems to me that would be an area that would be ripe for research-in-body microsensors or nanosensors.

Mark: That's absolutely correct. As our technology improves and becomes smaller and smaller, we're finding that we can do invasive procedures with humans that were never done before. A good example is ultrasound. A small ultrasonic probe can be placed into the arteries of the heart and we can look at the thickness of the artery, look for atherosclerosis or inflammation of the vessel.

Technology is bringing the physician into the human body in a less invasive manner. There's no reason to think we can't do something with some type of sensing device that would look for glucose, for example. Diabetics can place needles into the skin and a pump controls the rate at which insulin is delivered to the body to regulate glucose levels. In the future, a small detector could be placed inside the body to detect changes in glucose. It can then send a signal to a pump to release more insulin or vice versa. Machines that simultaneously measure and administer insulin today are large and cumbersome. I think with the advent of microtechnology, we're going to see something like this in the near future.

What do you two see for the near future and the far future, say, 50 years down the line?

Paul: We've seen in the last 20 years the growth of huge semiconductor giants like Intel and Motorola, who largely aim their products toward consumer electronics. It's feasible that new industry giants may arise in the next 10 or 20 years. These companies may make complete sets of specialized DNA chips in very large quantities that would target the health care industry. The diagnostic equipment could be inexpensive, highly manufacturable, widespread, high-speed, and accurate.

Mark: In 50 years, I wouldn't be surprised if a paramedic arrives at the scene of an accident and attaches some kind of device, a sort of Star Trek medical scanner, to the patient to get multiple readings-temperature, blood pressure, serum glucose level, etc. It would be some kind of sensing device that would determine the state of stress the body is in. By determining what's going on right away, the paramedic can call into the emergency room and get specific instructions and carry out those instructions quickly for, say, someone having a heart attack, a trauma victim, or someone having a diabetic reaction.

Paul: From the national defense point of view, people are talking about the electronic battlefield where each soldier has strapped on to him or herself a device that monitors blood levels and is able to tell what kind of condition the person is in during the heat of battle. Such information can then be relayed back to some control center where command officers can then know where everybody is and what condition they're in. This "battlefield telemedicine" could greatly influence the outcome of a battle.

Also, field medicine could be enhanced by microchip dogtags with medical histories, which can be telecommunicated to medical experts at remote sites.

Talk about lack of privacy.?

Mark: They do that when astronauts go into space. They can read their EKGs. The astronauts have different sensors on them to tell Mission Control what's going on. The neat thing with the newer technology is that such sensors will be precise, accurate, small, and very convenient to a person who's wearing it.

Paul: Such a sensor would have been very valuable in the Persian Gulf War. Some recent reports suggest that many thousands of our troops were exposed to chemical gases. Sensors might have been able to avoid that situation.

Mark: With the new technology, the level of sensitivity is going to increase logarithmically. It will be very sensitive.

Anything else?

Paul: It's a new and very exciting field. There's a lot of room for growth. We expect to see a lot of new jobs and new kinds of industries. So, it's a very exciting time.

Paul L. Gourley is a distinguished member of the technical staff at Sandia National Labs in the Nanostructure and Semiconductor Physics Department. He is a fellow of the American Physical Society, member of OSA and SPIE, chairman of the Photonics West '97 conference on Microfabrication for Biomedical Applications, and serves on the advisory board to Micro-Optical Devices, Inc. He received two DOE Basic Energy Sciences awards for outstanding research in semiconductor photonics in 1985 and 1993, has published more than 120 papers and one book, and holds several U.S. patents. He has made pioneering contributions in the areas of excitons in semiconductors, strained-layer semiconductor materials, vertical cavity surface-emitting lasers, and microcavity laser cytometry. His current interests include biomedical applications of semiconductors and lasers, semiconductor microlasers and broadband emitters, and solar UV protection for skin and eyes.

Mark Gourley, M.D., is Attending Rheumatologist at the Washington Hospital Center, Washington, D.C. Prior to joing the Hospital Center, he performed basic and clinical research at the National Institutes of Health, Bethesda, Maryland. His research investigations include the study of molecular and immunologic derangement in patients with autoimmune illnesses such as systemic lupus erythematosus. He performed clinical research investigating new therapies for patients with autoimmune diseases and studied the epidemiology of osteoporosis in patients with systemic lupus erythematosus. He is a member of the American College of Rheumatology and has many publications in both basic science and clinical journals describing disease pathophysiology and therapy of lupus.

They were interviewed by Frederick Su.