Gaia Vince and Clare Wilson

A man lies comatose on an operating table. The enormous spider that hangs above him has plunged four appendages into his belly. The spider, made of white steel, probes around inside the man's abdomen, then withdraws one of its arms. Held in the machine's claw is a neatly sealed bag containing a scrap of bloody tissue.

This is a da Vinci robot. It has allowed a surgeon, sitting at a control desk, to remove the patient's prostate gland in a manner that has several advantages over conventional methods. Yet the future of robotic surgery may lie not only with these hulking beasts but also with devices at the other end of the size spectrum.

The surgeons of tomorrow will include tiny robots that enter our bodies and do their work from the inside, with no need to open patients up or knock them out. While nanobots that swim through the blood are still in the realm of fantasy, several groups are developing devices a few millimeters in size. The first generation of "mini-medibots" may infiltrate our bodies through our ears, eyes and lungs, to deliver drugs, take tissue samples or install medical devices.

The engineering challenges are formidable, including developing new methods of propulsion and power supply. Nevertheless, the first prototypes are already being tested in animals and could move into tests on people in the not-too-distant future. "It's not impossible to think of this happening in five years," says Brad Nelson, a roboticist at the Swiss Federal Institute of Technology (EHT) in Zurich. "I'm convinced it's going to get there."

It was the 1970s that saw the arrival of minimally invasive surgery -- or keyhole surgery as it is also known. Instead of cutting open the body with large incisions, surgical tools are inserted through holes as small as 1 centimeter in diameter and controlled with external handles. Operations from stomach bypass to gall bladder removal are now done this way, reducing blood loss, pain and recovery time.

Combining keyhole surgery with the da Vinci system means the surgeon no longer handles the instruments directly, but via a computer console. This allows greater precision, as large hand gestures can be scaled down to small instrument movements, and any hand tremor is eliminated. There are over 1,000 da Vincis being used in clinics around the world.


There are several ways that such robotic surgery may be further enhanced. Various articulated, snake-like tools are being developed to access hard-to-reach areas. One such device, the "i-Snake," is controlled by a vision-tracking device worn over the surgeon's eyes. It should be ready for testing on patients within four years, says developer Guang-Zhong Yang, a roboticist at Imperial College London.

With further advances in miniaturization, the opportunities grow for getting medical devices inside the body in novel ways. One miniature device that's already been tried and tested is a camera in a capsule small enough to be swallowed.

In conventional endoscopy, a camera on the end of a flexible tube is inserted either through the mouth or the rectum, but this does not allow it to reach the middle part of the gut. The 25-millimeter-long capsule camera, on the other hand, can observe the entire gut on its journey. More sophisticated versions are being developed that can also release drugs and take samples.

The capsule camera has no need to propel itself because it's pushed along by the normal muscle contractions of the gut. For devices used elsewhere in the body, some of the key challenges are developing new mechanisms for propulsion and power supply on a miniature scale.

One solution is to have wires connecting the robot to a control unit that remains on the outside of the body. This is the case for a robot being developed for heart surgery, called HeartLander.

Operating on the heart has always presented enormous challenges, says Marco Zenati, a heart surgeon at the University of Pittsburgh, Pennsylvania, who's one of the device's inventors. Conventionally the heart is stopped and the patient hooked up to a heart-lung machine. A more recent approach is to perform keyhole surgery on the beating heart, but even so several incisions must be made, and the left lung must be partly deflated to allow access, requiring a general anaesthetic.

The HeartLander robot is designed to be delivered to the heart through a single keyhole incision, from where it can crawl to the right spot. The heart does not have to be stopped, and the left lung need not be deflated, so the patient could be breathing naturally, with just a local anesthetic. "Coronary surgery can become an outpatient procedure," says Cameron Riviere, the team's roboticist, based at Carnegie Mellon University in Pittsburgh.


The 20-millimeter-long HeartLander has front and rear foot-pads with suckers on the bottom, which allow it to inch along like a caterpillar. The surgeon watches the device with X-ray video or a magnetic tracker and controls it with a joystick. Alternatively, the device can navigate its own path to a spot chosen by the surgeon.

The HeartLander has several possible uses. It can be fitted with a needle attachment to take tissue samples, for example, or used to inject stem cells or gene therapies directly into heart muscle. There are several such agents in development, designed to promote the regrowth of muscle or blood vessels after a heart attack. The team is testing the device on pigs and has so far shown it can crawl over a beating heart to inject a marker dye at a target site (Innovations).

Another use would be to deliver pacemaker electrodes for a procedure called cardiac resynchronisation therapy, when the heart needs help in coordinating its rhythm. At the moment, the electrodes are delivered to the heart by pushing them in through a vein. Riviere's group is devising electrodes that the HeartLander could attach to the outer surface of the heart. They have tested this approach successfully on one live pig, and expect to start trials in people in about four years. Riviere says there is growing evidence to show that the technique works best when the electrodes are sited in certain areas that are hard to access from inside the veins. "The HeartLander can crawl around to the best position," he notes.

While the HeartLander could in theory be used in other parts of the body, in its current incarnation it has to be introduced through a keyhole incision thanks to its size and because it trails wires to the external control box. Not so for smaller robots under wireless control.

One such device in development is 5 millimeters long and just 1 millimeter in diameter, with 16 vibrating legs. Early versions of the "ViRob" had on-board power, but the developers decided that made it too bulky. Now it is powered externally, by a nearby electromagnet whose field fluctuates about 100 times a second, causing the legs to flick back and forth. The legs on the left and right sides respond best to different frequencies, so the robot can be steered by adjusting the frequency.

ViRob's developers at the Technion-Israel Institute of Technology in Haifa, are investigating several applications including taking tissue samples, delivering cancer drugs and getting a camera to hard-to-reach areas, such as deep within the lungs. The size of the camera is a limiting factor -- the smallest models in development are 1.5 millimetres in diameter -- but cameras get smaller every year, notes engineer Moshe Shoham.

The team would like their device to operate inside large blood vessels, but it is not yet powerful enough to withstand blood flow. "We don't want it swept away," says Shoham.

The first application for ViRob may benefit people born with hydrocephaly -- fluid on the brain -- as it may be able to extend the life of the shunts placed in the brain to drain the excess fluid. Over time such shunts tend to get blocked, and so need replacing every five to 10 years, entailing major brain surgery. Shoham says a self-cleaning shunt could be made by installing a ViRob permanently inside. About once a month it would be activated to send the device scuttling up and down the shunt, which patients might be able to do at home.

Another possible application might aid the insertion of cochlear implants. Used by deaf people, these are small electrodes placed within the delicate spiral-shaped cochlea to stimulate the auditory nerve. Shoham says ViRob would be able to carry the implant deeper inside the cochlea than can currently be done, giving patients better hearing. "The further you go into the cochlea, the more cells you excite," Shoham explains.

He reckons that tests on people are just a couple of years away. His team has a proven track record, having already commercialized a robot the size of a soft-drink can for a type of spinal surgery that involves fusing two vertebrae together. Called SpineAssist, the device is clamped over a keyhole incision on the spine, through which it finds the right spots on the vertebrae for the screws.

While the ViRob can crawl through tubes or over surfaces, it cannot swim. For that, the Israeli team are designing another device, called SwiMicRob, which is slightly larger than ViRob at 10 millimeters long and 3 millimeters in diameter. Powered by an on-board motor, the device has two tails that twirl like bacteria's flagella. SwiMicRob may one day be used inside fluid-filled spaces such those within the spine, although it is at an earlier stage of development than ViRob.

Another group has managed to shrink a medibot significantly further -- down to 0.9 millimeters by 0.3 millimeters -- by stripping out all propulsion and steering mechanisms. It is pulled around by electromagnets outside the body. The device itself is a metal shell shaped like a finned American football and it has a spike on the end.

The developers at ETH Zurich are focusing on eye surgery because it requires such a high level of precision -- hand tremor can be a major problem for surgeons operating here. The other draw is that this medibot's progress inside the eye can be monitored by viewing the eye through a microscope.

One application for the ophthalmic robot, as they call it, is to measure oxygen levels at the surface of the retina, an indication of its blood supply. For this, the shell is coated with a photoluminescent chemical, the brightness of which depends on oxygen concentration.

The device could also be used to treat a major cause of blindness known as retinal vein occlusion, which occurs when a blood clot blocks the major vein at the back of the eye. Various drugs are being investigated as treatments, such as one that dissolves blood clots, but they are hard to deliver. At the moment a kind of access port known as a trocar is placed into the surface of the eye, and a needle is inserted to inject the drug into the vein, but getting the needle to the hair-thin blood vessel demands great surgical skill.

Once the ophthalmic robot is delivered through the trocar, on the other hand, it can be guided to the blocked vein by its magnetic propulsion system. Its spike pierces the blood vessel, and the drug, which coats the device, diffuses into the vein.

The Swiss team is experimenting with even tinier versions of the device that fit inside the barrel of a needle and would simply be injected into the eyeball, avoiding the need for a trocar.

"We can make these smaller, but if we make them too small they cannot exert enough force to penetrate a vein," says Nelson.

Another refinement, he says, would be to make a biodegradable device that would not have to be removed from the eye. The shell would be made from a polymer, with an embedded metal particle to respond to the electromagnets. Once the polymer dissolved, the metal particle would be absorbed into the bloodstream and eventually excreted.

The team has been testing its devices on eyes removed from butchered pigs, and also on those of chicken embryos incubated in a Petri dish -- a set-up that eye surgeons often practise on. So far, they have shown that the robot can be put into the birds' eyes, steered to the right place and pierce the retinal vein.

The Swiss team is also among several groups who are trying to develop medibots at a vastly smaller scale, just nanometers in size, but these are at a much earlier development stage. Shrinking to this scale brings a host of new challenges, and it is likely to be some time before these kinds of devices reach the clinic.

Nelson hopes that if millimeter-sized devices such as his ophthalmic robot prove their worth, they will attract more funding to kick-start nanometer-scale research. "If we can show small devices that do something useful, hopefully that will convince people that it's not just science fiction."


© Gaia Vince and Clare Wilson, New Scientist Magazine






Rise of Miniature Medical Robots: Fantasy Fast Becoming Reality