快猫短视频

ACCORDING to ancient history, the Persian warrior Hegesistratus escaped from his enemies in 484 BC by cutting off his own foot to escape his shackles. The soldier later took to wearing a wooden foot. If the tale is true, this would be the first recorded use of a prosthetic limb.

Artificial legs, arms and hands 鈥� often hooks 鈥� followed. But it took two millennia for the next major advance to emerge, when French army surgeon Ambroise Pare designed a more natural 鈥渁rticulated鈥�, or jointed, leg in 1529. The prosthetic leg had made the technological leap from rigid pole to a more natural, flexible limb.

Today we are on the verge of seeing an even more important advance in prosthetics that should transform the lives of people who have lost arms or legs 鈥� the development of the ultimate bionic limb, wired directly into the nervous system. Recipients will be able to move these devices by the power of thought alone, as well as receive lifelike sensations from them. Sensory feedback from artificial limbs is in fact turning out to play a surprisingly important role in the development of responsive, naturalistic devices.

As ambitious as these goals may sound, the debut of the bionic limb could be surprisingly close 鈥� perhaps within the next five years. Researchers in the US, Europe and Japan have overcome many of the obstacles, and human trials of the first crude prototypes have already begun. 鈥淎 large part of it is now just putting the pieces together,鈥� says Ronald Riso, a biomedical engineer at Aalborg University in Denmark who is one of the prosthetic pioneers.

Today鈥檚 artificial limbs are already highly advanced. The most sophisticated prosthetic hands can be made to open and close via electromyography, a recording technique where special sensors detect electrical activity within the wearer鈥檚 arm muscles, which are connected by circuits to the hand鈥檚 motors. So by flexing their arm muscles, wearers can make their prosthetic hand grip objects or release them. The motors are powered by rechargeable batteries, which last for about a day.

While these devices are no doubt greatly valued by some amputees, they can be very awkward to use, especially initially. Some studies have shown that as many as half of owners do not use their prosthetic limb regularly because the device is ineffective, cumbersome, or just plain ugly.

So what will the artificial limbs of the future be like? They may well look something like Cyberhand, a prosthetic hand that will be jacked directly into the nervous system. It is being developed by a collaboration of scientists from four European countries.

For a start, the hand鈥檚 mechanical design is going to be more sophisticated. Unlike existing prosthetics, Cyberhand鈥檚 five digits will each have independent motors that can be controlled separately. According to Paolo Dario, a mechanical engineer at the Sant鈥橝nna School of Advanced Studies in Pisa, Italy, who is leading the team developing the hand鈥檚 machinery, Cyberhand users should be sufficiently dextrous to use a pair of scissors.

Fitting all the joints, motors, sensors, electronics and batteries into such a small space is a significant engineering challenge. Weight is another limiting factor, says team member Fabrizio Vecchi, as the hand should ideally be no heavier than about 600 grams, like a real one. 鈥淭he number of batteries could be a problem,鈥� he points out, given all those power-hungry motors. But Vecchi reckons that by the end of the year, a prototype Cyberhand will be ready that can be controlled by electromyography, like existing devices.

Meanwhile, other Cyberhand researchers are working on ways to connect the hand to the same nerves in the forearm that once served the person鈥檚 original hand. To do so they are having to design new kinds of electrodes that can act as an interface between the hand鈥檚 electronic circuits and the body鈥檚 nerves.

One set of electrodes will need to record the outgoing (or 鈥渕otor鈥�) nerve signals coming from the brain, and translate these into electronic signals to operate the hand鈥檚 machinery. A second set must translate the artificial sensations generated within the hand into signals that can be picked up by the sensory nerves leading from the arm to the brain.

Why should sensory feedback be so important? We couldn鈥檛 use our real limbs very well without it. Imagine trying to pick up a glass when you don鈥檛 know how hard you are gripping it 鈥� too lightly and it will slip, too hard and it could be crushed. As for reaching into your pocket for some keys, or fumbling for the light switch in the dark, sensation is vital.

There are four different types of touch receptors in the skin, sensing acute pressure, sustained pressure, vibration or skin stretch. But there鈥檚 another kind of sensory perception that also plays a crucial role 鈥� the 鈥減roprioception鈥� signalling from our muscles, tendons and joints, which gives the brain information on the body鈥檚 positioning and movement. This is vital for precise control of movements and general body awareness 鈥� to let you know your arm鈥檚 position without having to look at it, say.

快猫短视频s believe such sensory feedback will play an important role in helping the brain learn to use an artificial limb 鈥� just as a baby learns to use its limbs for the first time. When sensory neurons convey the information that a body part is moving in the desired way, that strengthens connections between the neurons that brought about that movement.

Over the past few years, novel animal research by Miguel Nicolelis and colleagues at Duke University in Durham, North Carolina, has lent credence to this idea. They managed to train monkeys to move a mechanical arm just by willing it, using electrodes in the animals鈥� brains to detect their intentions (快猫短视频, 23 February 2002, p 26). Key to the project鈥檚 success, says Nicolelis, was allowing the animals to watch the arm鈥檚 movement, to give them positive feedback when they were doing it right. 鈥淲e noticed a reorganisation of the cortex,鈥� he says. 鈥淭he properties of the robot arm were being assimilated as if they were a part of the animal鈥檚 own body.鈥� Nicolelis believes that if the animals had been given proprioception and touch sensations from the arm, they would have learned even faster.

Well wired

So what are the challenges involved in wiring a prosthetic hand directly into the nervous system? A major hurdle is devising electrodes that can communicate with individual neurons and that will stay in place. 鈥淚t is very difficult to fix the microelectrode at a certain position for a long time,鈥� says Mabuchi Kunihiko, a medical engineer heading a similar project to Cyberhand, based at the University of Tokyo.

The state-of-the-art 鈥渃uff electrode鈥� is a collar-shaped device that encircles whole nerve fibres, comprising several thousand neurons. Riso鈥檚 Cyberhand team at Aalborg University has tried temporarily implanting these into volunteers. The patients reported it felt 鈥渢ingly鈥�, says Riso, a little like a vibration, most probably because groups of neurons were being stimulated simultaneously. While this is not ideal, it鈥檚 better than nothing, he says.

The team is now recruiting volunteers who are willing to have the electrodes permanently implanted to assess how acceptable these sensations are in the long term. The patients will not initially get a prosthetic hand, but the researchers plan to use this group as guinea pigs for future versions of Cyberhand. The patients should not have wires sticking out of their stumps, fortunately, as it is possible to use radio signals to communicate between the hand and implanted electrodes.

Another Cyberhand research group, led by Xavier Navarro, a neuroscientist at the Independent University of Barcelona, Spain, is trying a different approach. They are investigating using a 鈥渟ieve electrode鈥�, currently used in neuroregeneration research. As the name suggests, each electrode has many holes in it, each one capable of delivering a separate electrical signal. The aim is to get individual neurons to grow through the holes, so they can be stimulated one at a time.

Unfortunately, peripheral neurons are a gregarious lot 鈥� they seem to prefer to grow alongside other neurons. In experiments on rats, Navarro鈥檚 team has so far managed to get groups of between 10 and 40 neurons to grow through holes with a diameter of 40 micrometres. The trouble is, when he makes the holes any smaller, no neurons will grow through them. But 40 micrometres may do to start with, he says, so his group is now assessing the long-term safety of this approach with a view to starting human trials.

Perhaps the furthest advances in this field, however, come from researchers at the University of Utah in Salt Lake City, led by bioengineer Ken Horch. They are using a new type of electrode, nicknamed the 鈥渂ed of nails鈥�. This is a platform of needle-like points that each transmit separate electrical signals. The device is implanted so that each 鈥渘ail鈥� lies next to an individual nerve fibre. 鈥淭hey are fairly resistant to movement,鈥� says Horch.

Last year the researchers gained approval to implant this type of electrode into human amputees for a few hours. In as yet unpublished work, the electrodes were wired up to a table-mounted mechanical arm that had pressure sensors on its surface and proprioception sensors within the motors. According to Horch, when the researchers moved the arm or made it grip objects, the volunteers reported that they could indeed feel appropriate sensations. They even gave a rough indication of the amount of gripping force they were applying. 鈥淲e elicited touch and proprioceptive sensations of finger movement and position,鈥� he says. 鈥淎ccording to the subjects, the sensations felt normal.鈥�

Although this research is still in the early stages, the fact that the patients described natural-feeling sensations bodes well, and suggests that Horch鈥檚 electrodes are achieving close to one-to-one correspondence from electrode to neuron.

As if the prospect of tapping into a patient鈥檚 peripheral nervous system isn鈥檛 futuristic enough, other scientists are considering wiring one directly into the brain. Again, two sets of electrodes would be needed, to record outgoing motor signals, and pass on incoming sensory information. These would have to be relayed to the artificial limb either via electronic circuitry or radio.

Although this approach would clearly be more invasive, it has a distinct advantage, according to one of the researchers, John Chapin at the State University of New York. Brain implants would be more likely to stay put, because there wouldn鈥檛 be the movement of muscle and bone to jostle them about.

Work such as Nicolelis鈥檚 monkey research has shown that, in animals at least, it is possible to pick up the motor signals from the brain that control limb movement. Chapin is now collaborating with Nicolelis to try to improve the system by providing it with sensory feedback, which they are doing by implanting electrodes in the monkeys鈥� brains. They are about to start the extensive animal safety testing that will have to be carried out before such an invasive procedure can be carried out in humans.

Some researchers are trying to use this type of technology to benefit quite a different group of patients: those who have been paralysed after spinal cord injuries (see 鈥淕etting to grips with paralysis鈥�). They want to reconnect people鈥檚 brains to their limbs electronically, essentially bypassing the damaged spinal cord.

The first generation of bionic limbs will not, of course, be able to replicate the exquisitely precise sensory and motor functions of the flesh-and-blood limbs they are replacing. The first hands are likely to have only tens of sensors 鈥� rather less than the thousands in a human hand. For one thing, there is little point in putting lots of sensors in the device until we know how the brain will interpret data from an artificial limb. Will people eventually become so accustomed to the artificial sensations that they feel completely natural?

In a recent interview, paralysed actor Christopher Reeve spoke of how he missed being able to feel the warmth of someone holding his hand. 鈥淭his is what we鈥檙e up against,鈥� says Chapin. 鈥淚 don鈥檛 think we鈥檒l be able to restore all of the complexities of touch, but I think we can restore part of it. And that will be a big quality-of-life improvement.鈥�