Medical Microbots Take a Fantastic Voyage Into Reality

Engineers explore ways to take robotics to the limits of size and function

In the 1966 film Fantastic Voyage, scientists at a U.S. laboratory shrink a submarine called Proteus and its human crew to microscopic size and then inject the vessel into an ailing scientist. Once inside, Proteus motors its way through the bloodstream and into the brain, where members of the crew don scuba gear and use a laser gun to perform delicate surgery.

From our comfortable 21st-century perch, there is a lot in Fantastic Voyage to smile about. But the notion of performing medical procedures at microscopic scales is now slowly sneaking out of the realm of science fiction. Thanks to developments in microfabrication and other areas, researchers are pushing the limits on the size and capabilities of objects small enough to move through the human body.

In the past 10 years or so, a menagerie of whimsical-sounding designs has emerged: microrobots driven by bull sperm and bacteria, starfishlike microgrippers that can close their arms around tissue as they get warm, spinning magnetic helices that can deliver DNA to cells, steerable magnetic spheres packed with drugs, micromotors powered by gastric acid, and microscallops that can flap their way through the vitreous humor of the eye.

human os iconMany of these devices are still little more than laboratory curiosities, but others are being tested in animals. And some engineers are confident that tiny, untethered instruments will one day be used in medicine. “Our biggest impact will be in health care,” says microroboticist Metin Sitti, who leads the Physical Intelligence Department of the Max Planck Institute for Intelligent Systems, in Stuttgart, Germany. He coauthored a recent survey of the field for Proceedings of the IEEE.

With the right design, researchers say, a microrobot—or a swarm of them—could deliver a highly targeted dose of drugs or radioactive seeds, clear a blood clot, perform a tissue biopsy, or even build a scaffold on which new cells could grow.

These sorts of activities could help extend two current trends in medicine: diagnosing diseases earlier and targeting therapies more precisely, says Bradley Nelson, a professor of robotics and intelligent systems at ETH Zurich. “The dream is the Fantastic Voyage,” he says.

Realizing that dream will mean overcoming a range of engineering obstacles. At microscopic scales, nearly every aspect of robotic operation needs to be rethought; power and movement become especially tricky. And working in the human body applies extra constraints: You need to be able to keep track of where an object is, make sure it isn’t toxic and won’t injure tissue, and design it to degrade safely or leave the body once its mission is complete.

“It’s just been in the last few years that I think the community has started to get a handle on these fundamental problems,” Nelson says. Now, he adds, the focus has turned to what can be done with the technologies researchers have in hand.

Footage Courtesy of Johns Hopkins University

Medicine is already embracing miniaturization, and some technology now makes its way through the human body without any tether to the outside world. There are, for example, battery-powered gadgets about the size of a large vitamin pill that can snap images of the esophagus, intestine, and colon as they move along.

And in 2012, the U.S. Food and Drug Administration gave Proteus Digital Health, headquartered in Redwood City, Calif., the green light to market a much smaller swallowable technology: a single-square-millimeter silicon circuit that can be embedded inside a pharmaceutical pill.

“It’s the smallest ingestible computer in the world,” says Markus Christen, a senior vice president at Proteus. He is quick to note that its capabilities as a computer are limited. The Proteus chip carries neither an antenna nor a power source. Instead, it contains two electrode materials that are electrically connected when the pill surrounding the chip dissolves and the circuitry comes into contact with the stomach’s gastric juice. For 5 or 10 minutes, the chip has enough power—between 1 and 10 milliwatts—to modulate a current, transmitting a unique identifier code that can be picked up by an external skin patch.

Sizewise, the Proteus chip is just at the upper end of what might be considered a microrobot. The chip is more than sufficient, Christen says, to help patients keep track of drug consumption and help pharmaceutical companies monitor how closely subjects in clinical trials follow a regimen when they’re testing a new drug.

Making objects smaller yet more capable will require creative solutions. One of the biggest hurdles is power. Miniaturization doesn’t favor traditional chemical battery technology, says Max Planck’s Sitti. Shrink an object down below 1 millimeter, he says, and “the battery’s capacity will go down drastically.”

One alternative is wireless power transfer—piping in radio waves, for example, from outside the body to generate electricity. But this approach also becomes difficult at small scales. To harvest the energy, a microrobot would need some sort of antenna, and that antenna can’t be too small if it’s to collect a meaningful amount of energy. It must also stay fairly close to the source.

Given these limitations, engineers are looking at new ways to gather power for activities such as propulsion. One option is to create what are essentially small chemical rockets—objects that can react with substances in the body, such as gastric acid, to move around. Researchers are also exploring what can be done with biohybrids, in which bacteria are harnessed to do the swimming and perhaps even pursue a target based on signals, such as the change in concentration of a particular molecule.

In some cases, it may even be possible to do without any onboard source of energy. At Johns Hopkins University, in Baltimore, David Gracias and his colleagues have developed microgrippers—star-shaped devices that can measure less than 500 micrometers from tip to tip. The grippers can be made of materials that respond to environmental factors such as temperature, pH, and even enzymes. A temperature-sensitive gripper’s arms will close when exposed to the body’s heat. If placed well, the arms will close around tissue, performing a miniature biopsy.

Such grippers might provide a less invasive way to screen for colon cancer in patients who suffer from chronic inflammatory bowel diseases. Today, Gracias says, such screening can involve taking dozens of samples with forceps, in an effort to get good statistical coverage of the interior surface of the colon. Instead, a doctor could insert hundreds or thousands of microgrippers into the colon through a tube and then retrieve them using a magnet or, later, by sifting through the patient’s stool.

Based on tests in live pigs, Gracias’s team estimates that about one-third of the grippers capture tissue. Others may come up empty-handed because they have the wrong orientation or close before reaching anything. But he says this approach, which minimizes the cost and maximizes the ease of manufacturing, could be powerful.

“The typical idea has been that you have one device that you guide precisely [to perform a] surgical procedure,” Gracias says. His strategy borrows a page from the imperfect world of biology: “If you have a large number of not-perfect devices, you may be able to achieve the same functionality as one perfect one.”

The gastrointestinal tract is a fairly forgiving place to work inside the human body. It’s relatively large and easy to access externally, and it automatically funnels objects through the body. Exploring trickier locations, such as the eye, the brain, and the bloodstream, will likely require more sophisticated microrobot designs.

One significant hurdle is the machines’ potential to trigger clots. “When you talk to clinicians, one thing that makes them go white and never want to talk to you again is any kind of notion of putting something solid in the bloodstream,” says John Rogers, a pioneer of soft electronics for the body at the University of Illinois at Urbana-Champaign. “There are just really serious consequences of any kind of structure that’s free-floating and just traveling around.” [For more about Rogers and his work, see “A Temporary Tattoo That Senses Through Your Skin,” in this issue.]

Precise placement of microbots is therefore crucial. Even the most sophisticated microswimmers, ones capable of following a change in pH or temperature, might not be able to combat the powerful currents in the bloodstream. “The reality is, these things are not going to swim for long distances in your body,” says ETH Zurich’s Nelson. An autonomous swimmer might be able to muster only 20µm or so of fast, directed motion, he says, so it’s likely that external guidance will be needed to get the device most of the way to its destination.

One of Nelson’s targets is the retina. Today, drugs designed to treat the retina can be injected into the eye, where they slowly diffuse, but only a fraction of such a dose may reach its target. Microbots laden with drugs, Nelson says, could potentially deliver them in a more targeted manner, reducing doses as well as side effects.

One obvious strategy for guiding a robot to the right spot is to build it out of magnetic materials and then steer it externally with magnets. Researchers have used MRI machines to do this in animals. But Nelson, Sitti, and others are pursuing less powerful electromagnet configurations capable of even greater control.

Moving a microrobot with magnetic fields turns out to be surprisingly tricky. “We’re still learning about the mathematics and physics of that,” Nelson says. To move an object with a robotic arm in any arbitrary way, he explains, you need six actuators for a full six degrees of freedom: movement in the x, y, and z directions, and rotation around each of those axes. When he and his colleagues worked out a way to finely control a simple magnetic microrobot with five degrees of freedom, they found that eight separate external magnetic coils were needed. Adding in the sixth degree, Nelson says, requires that the microrobot have a more complex magnetic profile than a simple bar magnet.

Nelson’s team can use its magnets to control a helical microrobot with magnetic fields of less than 10 millitesla, a fraction of what’s created in an MRI. “We can twist these helices and cause them to corkscrew and move forward,” he says, much in the same way that E. coli bacteria propel themselves by rotating their flagella. Earlier this year, his team reported that it had successfully used coated versions of these artificial bacterial flagella in the laboratory to deliver genetic material to human cells.

In the near term, Nelson is looking to see how his magnetic control technology could be used by doctors in a tethered fashion, as a way of finely guiding the tips of catheters through the cardiovascular system. But in the long term, he’s exploring what might be done once this physical connection to the outside world is severed. For him and many other researchers, the possibilities seem vast—as big as the human body itself weighed against the complex, ever-moving world of the microscale.

This article originally appeared in print as “Microbots on a Fantastic Voyage.”