A deaf woman is talking to me on the phone. Her name is Joanne Syrja, and she is explaining to me how her hearing deteriorated over time. She was never able to hear high frequencies, she says, and as the years went by, that ceiling of sound descended until she could hear nothing at all. But she can now hear me.
The questions I ask her are transformed into pulses of electricity that travel over telephone lines and are changed back into waves of sound at her receiver. Then a microphone sitting in her ear transforms my voice once again into electricity. In the fall of 1991, when Syrja was 44, surgeons put a device known as a cochlear implant deep in her ear. It uses the microphone’s signals to stimulate her auditory nerve endings, and her brain senses the pattern of that stimulation as my voice.
I ask her what surprises her implant has given her. It was last March, she says. I was going into my office, and there are some trees right in front of the doorway. I heard a noise, and I saw this bird sitting there. Every time its mouth opened, I heard this noise. I never heard birds before in my life. It was incredible. I just stood there with tears running down my face. I knew what I heard was the bird chirping, and it was beautiful.
Later I speak over the phone with Roslyn Rosen. Like the vast majority of deaf people, Rosen was deaf before she could learn how to speak. She is now the dean of continuing education at Gallaudet University- -a unique college for the deaf in Washington, D.C.--and president of the National Association of the Deaf. I’m not actually speaking to her--an American Sign Language interpreter translates for us. The translation is so fast and smooth that I occasionally have to remind myself that I’m not talking directly to Rosen. She is blasting the FDA’s 1990 approval of cochlear implantation in children who are born deaf. Most of the cochlear implant children are still deaf children, Rosen says. And at $20,000 to $40,000 for the procedure, the real losers are the child, the family, and the insurance companies.
It has come to this: artificial senses, once the technological property only of fantasy and fiction, are now triggering passionate debates on the propriety of their use. It seems safe to predict that the cochlear implant will be joined in the coming decades by other devices that will provide some kind of artificial sight and touch. How they will change people’s lives, however, cannot be predicted.
Twenty years ago Robert Schindler, then a medical resident at the University of California at San Francisco, discovered that a few researchers thought a machine could make deaf people hear. Their reasoning was seductively simple.
When sound waves enter a person’s ear, they push insistently against the eardrum. That vibration gets passed on through a group of hinged bones, the last of which taps against the opening to the liquid- filled spiral chamber known as the cochlea. Waves roll through it, causing the delicate bristles lining its interior to bend. As they bend, pores on their surface get pried open, allowing charged atoms floating in the liquid to flow in. This creates an electric signal that travels to nerve endings in the cochlea and on into the brain. In the process, the cochlea manages to separate complex sounds, like speech, into their component frequencies. Different frequencies make the bristles in different spots in the cochlea quiver strongly. The higher the note, the closer to the entrance of the cochlea this sensitivity appears.
The most common cause of deafness is the death of the cells bearing those sensitive bristles--the hair cells. Even with the hair cells gone, though, the nerve endings often still work. A few researchers, Schindler discovered, were arguing that a machine could stimulate these nerve endings--and that if it stimulated them in different places in the cochlea, it could make the brain hear the appropriate frequencies.
Most experts doubted it. The orthodoxy of the day said this was impossible, says Schindler. It said that you’d kill the nerve, that there was no way you could represent the complexity of speech. All the implant wearers could ever hope to hear was a buzzing sound.
Schindler helped implant some of the earliest prototypes, though, and they showed promising results. I caught the bug and said, ‘This is going to be what I want to do for a lifetime.’ Over the following two decades he and other researchers discovered that, contrary to the reigning opinion of the day, cochlear implants did not kill the auditory nerve and that although some people could indeed get nothing more than a buzzing noise with them, others could hear again. Among postlingually deaf adults-- those who have gone deaf after learning to speak--cochlear implants have allowed about a fifth of the recipients to hear and speak as well as Syrja. Three-fifths can use the implant as an aid to lipreading. The other fifth get no real benefit.
The basic assembly of all cochlear implants is the same. The microphone in Joanne Syrja’s ear is connected to a sound processor that she carries on her belt. There sounds are manipulated so that important patterns are highlighted, thus making it easier for her to understand. The processor splits the sounds into bands of frequencies and sends them to a receiver implanted in Syrja’s skull. From there the signals travel down an insulated cable, which ends as electrode-studded wires snaked into the cochlea. Each electrode receives information about a single band of frequencies, and it stimulates the nerve endings around it in a matching pattern. The brain can then convert the pattern of stimulation into the range of frequencies that make up the sounds.
During the last decade Schindler and other implant researchers began to reap the rewards of their work. In 1984 the FDA approved the first commercial implants for adults, and in 1990 the agency extended their use to children. Now more than 7,000 people worldwide, over 2,000 of them children, wear cochlear implants. Researchers working on other artificial- sense devices, burdened with huge prototypes and hazy results, talk about the day when their machines will be as successful as the cochlear implant. This has been a classic scientific revolution, says Schindler.
Still, a surgical procedure with a 20 percent complete-success rate has room for improvement. The latest research, focusing on making the sound processor more sophisticated and speeding up the electrodes’ response to changes in sound, may make the devices more effective. Schindler, for one, believes the number of deaf people who will get major benefits from implants will rise dramatically, to perhaps as high as 75 percent.
For some 2,000 deaf Americans, however, cochlear implants are, and will continue to be, pointless. Their deafness is caused by a rare disease known as neurofibromatosis, in the course of which tumors may grow on the auditory nerve itself. When surgeons cut the tumors out, they can’t avoid cutting the nerve as well. For such a person to hear again, a device would have to tap into the hearing pathway beyond the nerve--in other words, into the brain itself.
Researchers at the House Ear Institute in Los Angeles have designed such a device. It consists of an electrode array that is placed on a part of the brain known as the cochlear nucleus. This is the first relay station inside the brain for signals coming in on the auditory nerve. By firing an electrode on the surface of the cochlear nucleus, the implant can give a sensation of sound, but nothing as clearly distinguished as speech. The implant has been moving smoothly through the FDA approval procedure, and it is likely to become the first commercially available brain implant that restores a sense.
For researchers trying to restore hearing, brain implants are a wild new frontier. For those involved in artificial vision, however, the frontier has been home for more than 25 years. Only about 20 percent of blind people still possess working optic nerves. Researchers trying to blaze a trail to artificial sight, therefore, have no choice but to take the most ambitious route.
Nerves carrying visual information run from the eyes to the back of the brain, where they plunge into the area known as the visual cortex. Here the signals are organized into maps that correspond spatially to the outside world--if someone loses a patch of his visual cortex, he becomes blind in a corresponding patch of his field of view.
Since the 1930s researchers have known that by stimulating the visual cortex they can make blind people see points of light. In the 1960s several groups began to investigate the possibilities of exploiting this phenomenon. In one study, at the University of Utah, blind volunteers agreed to have an array of electrodes placed on the surface of their visual cortex; connecting wires ran out through holes drilled in the skull to a jack on the scalp. When the researchers ran current through the electrodes, they stimulated the neurons in the visual cortex in such a way that the subjects saw an array of up to two dozen dots floating in the darkness before them.
To the more imaginative scientist, these results were wonderful. A blind person could wear a pair of eyeglasses with thumb-size cameras attached to them. The images captured by the cameras could be sent through a wire to a jack implanted in the back of the person’s head, and on to hundreds of electrodes inserted into the visual cortex. There the signals would create a pattern of dots that could form a televisionlike image of the world outside.
How good would the image have to be to be useful? Recently a team of Utah researchers found out by having volunteers with normal vision put on headgear that let them see what the world would look like through an artificial eye. A black-and-white video camera took in the view, which was broken down into different numbers of dots. The volunteers moved around a room and tried to open doors, and they tried to read. It turned out that if they could see 625 dots, they could move around without bumping into objects and read 150 words a minute.
For more than 20 years the National Institutes of Health has bankrolled a project known as the Neural Prosthesis Program, one of whose longest-running undertakings has been to develop the electrodes that could make possible an apparatus such as that envisioned by the Utah researchers. The problem with the original electrodes is that they sit on the surface of the visual cortex while the actual map is located about a tenth of an inch underneath it. And as the current released by the electrodes travels into the brain, it spreads out. By the time it reaches the visual map, the signals can interfere with each other. Several hundred such electrodes can’t create a crisp image. Instead they yield something like a television screen image in which each pixel--each dot--is the size of a dime and overlaps its neighbors. What the Neural Prosthesis Program wanted was an electrode like a needle, which could actually penetrate the brain to get close to the visual map.
It took more than 15 years to build it. Terry Hambrecht, the director of the program, keeps one in his Bethesda office to show visitors. The electrode looks like an eyelash, but one section of it, which would connect to leads leaving the skull, is made of gold, while the other part, which would penetrate the brain, is made of insulated iridium. A little blob of epoxy keeps the two sections together. At the tip of the iridium end is a microscopic bit of iridium oxide no larger than a neuron--only .001 inch across. When current runs through the wire, it enters the brain from only this single tip, keeping the electric field focused.
Hambrecht and his colleagues tested the device on a blind woman in November 1991. They implanted 38 electrodes in her head, and she reported seeing points of light corresponding to all but four of them. By asking her where each one appeared in her field of view, then shutting off some and turning on others, the researchers were able to make the points form simple shapes, such as the letter I. The researchers hope to have 250 electrodes ready to implant in another person by next January. With that many points of light, the implant wearer may even be able to recognize letters on a computer screen.
These electrodes are labors of love. Each has to be handcrafted, then carefully inserted into the brain by a neurosurgeon. Another project going on at the Neural Prosthesis Program, however, could lead to mass- produced, multielectrode probes. Bill Heetderks, who oversees the research, has his own electrode to show off. It has a barely visible square head of microelectronics and a long, slender shank. It was etched out of a wafer of silicon and has 16 contact sites, each of which can be independently controlled. Heetderks and his colleagues are getting ready to move up to 64 contacts; conceivably, they could go up to hundreds. As with computer chips, one wafer of silicon can cheaply yield many probes. Once you’ve got the basic design, you can churn them out like jelly beans, says Heetderks.
With these probes, surgeons wouldn’t have to carefully move a few precious electrodes until they found the precise spots where they would produce the best image. You just put the probe into the basic area, Hambrecht explains, and then you electronically scan the electrodes until you find what you want and forget about the other electrodes. We’re going to waste a lot of contacts, but who cares?
Not all the research into restoring sight involves brain implants. A fifth of blind people have optic nerves that work normally, with endings in the eye still intact. Their blindness is caused by genetic defects or diseases that destroy the photoreceptors that line their retina and convert light into electricity. All these people would need to see again is an artificial retina.
Two teams of researchers have been working to build a retina. One team is in the South, at Johns Hopkins, Duke, Research Triangle Park, and North Carolina State; the other is in the North, at MIT and Harvard. The artificial retina the northern group envisions incorporates the latest video-camera technology: light will strike a modified charge-coupled device, or CCD, which will convert the light into electric pulses. The southern group is using the newest solar cell technology for the same purpose. In both cases the device will be placed in front of the retina, will receive light on its front side, convert it into electric pulses on corresponding spots on its back, and stimulate the appropriate nerve endings.
Both teams have developed surgical techniques for getting the implant into the eye and fastened to the retina. The northern researchers are now putting blank chips into the eyes of rabbits to see how well a living eye tolerates having electronic hardware sitting inside it. By the end of the year they hope to put a prototype in one eye of a rabbit. They’ll measure electrical activity in the rabbit’s visual cortex and do behavioral tests to assess how well the animal can see.
At the very least, a retinal implant should be able to let a blind person see the outlines of objects around him, allowing him to walk without a cane. Eventually the artificial retinas might improve to the point where the implant wearer could read with them; beyond that, it’s anybody’s guess.
After hearing and vision, touch is the sense most likely to be mimicked by technology. In some ways, however, it presents a more complex problem, since it is so tangled up with other bodily functions--people who lose their sense of touch generally lose their ability to move as well. But advanced electrodes such the ones Heetderks is working on could someday help.
Surprisingly, the nerves in a paralyzed limb are often in perfect working order, sending fresh information toward the brain hundreds of times a second. Only when the signals reach a break like a severed spinal cord do they die out. But it’s possible to keep that information from being lost; it can be captured with a device known as a recording cuff electrode. The cuff is a silicone sleeve about an inch long. It is wrapped around a single nerve, and electrodes along its interior record all the currents that race past.
The first cuff was built 15 years ago by Andy Hoffer, now a neurophysiologist at Simon Fraser University in Burnaby, British Columbia. Hoffer originally developed the device to measure nervous activity in animals, but all along he’s been trying to refine its use for humans. Recently one of Hoffer’s graduate students, Morten Haugland, working with Thomas Sinkjær--a former Hoffer grad student--at Aalborg University in Denmark, found a way to put Hoffer’s ideas into action.
Strokes often cause lesions in the brain and can leave a person with only partial control of half the body. Walking becomes a great chore; because the person can no longer lift one foot, it just drags along the ground. A year ago Haugland and Sinkjær put a recording cuff on a foot nerve in a man who suffered from this condition. They also implanted a stimulating electrode on the front of his leg, just below the knee, and connected both devices to a microprocessor. Now when the man rolls his weight from the heel to the ball of his affected foot and the pressure makes the nerve fire, the recorder detects its signal. Once the microprocessor gets this message, it tells the stimulator to release a burst of electricity that makes the muscles around it contract. The man’s foot lifts, and he can walk comfortably.
Hoffer and Heetderks are looking far beyond dragging feet, however. Hoffer is trying to figure out how to put his cuffs on the nerves that monitor the position and angles of muscles and joints. These signals could help give a person a sense of where his limbs are in space. All this information could, in theory, be relayed to the appropriate part of a person’s brain via Heetderks’s electrodes. All a surgeon would have to do is put them in the part of the brain that contains the touch map of the body rather than the vision map. Heetderks also hopes someday to have recording electrodes that could be implanted in the motor control area of the brain and run stimulating electrodes in muscles. A patient could then feel the cup he was holding, think about moving it to his lips, and sense his arm bringing it there.
All this, though, Heetderks is quick to point out, is almost complete speculation. Neural-prosthetics researchers love to talk about how their devices might work in the far future, but as a rule they do everything they can to avoid being perceived as miracle workers. It will take years, if not decades, to bring these projects to fruition, and even then they won’t magically reproduce the full richness of a sense.
Despite the risk of raising expectations, the researchers work from the premise that what they are trying to do is intrinsically good. Others aren’t so sure. Are cochlear implants, for example, as valuable as they’re claimed to be? In a case like Syrja’s, the results are inarguably impressive. But she is not the future of this technology. Researchers and corporations are setting their sights on the far larger pool of prelingually deaf children, who became deaf before they learned to speak. How well cochlear implants do with children who cannot connect the sounds they hear with words they’ve heard before is unclear.
Harlan Lane, a psychologist at Northeastern University in Boston whose research includes testing postlingually deaf adult implant wearers, agrees that for these adults the technology does indeed show promise. But when the FDA approved implants for prelingually deaf children, he maintains, they blew it. They made the wrong decision scientifically, medically, and ethically.
Lane argues that the results of the preapproval trials and follow-up studies are not impressive--that only a few children in the trials could actually recognize randomly selected words without lipreading. And he points out that although the vast majority of deaf children are prelingually deaf, most studies lump them in with postlingually deaf children, and the results are thereby skewed. When Lane recently analyzed several such studies, he concluded that implants gave the prelingually deaf children almost no ability to recognize speech by ear alone. At best, Lane believes, it takes years of training for these children to recognize just a few spoken phrases, and with such a limitation they are unlikely to learn much in school. Learning a language early in life is crucial to understanding other subjects--as well as to simply thinking clearly. Yet despite much time devoted to training them, children with implants appear to be bound for a substandard understanding of spoken English.
Deaf children can more reliably learn American Sign Language, Lane says, yet there seems to be no particular interest among surgeons or implant-making corporations to make sure they do. Lane believes that the implanted children will end up trapped between two worlds: they can’t live the way hearing people can, and yet they won’t have grown up in the deaf community, using ASL. Unfortunately, he says, we’re going to have to wait 15 years until enough of them are damaged and enough research is done on the damage before we can persuade the FDA that they made a mistake.
Rosen, the Gallaudet University dean, believes that the advocates of cochlear implants don’t understand what it means to be deaf. Deaf people, she points out, have built themselves a language, a culture, and a history as legitimate as those of other minorities. Most people view deafness as a pathological condition and as a problem in search of a cure, says Rosen. We don’t see ourselves that way. We view ourselves as people who happen to not hear, and for whom life is still very good. Yet many involved with the development of cochlear implants, she says, believe that surgery and dim prospects for hearing are better than accepting oneself as deaf.
The debate over the worth of neural prosthetics isn’t limited to artificial hearing. James Gashel, the director of government affairs for the National Federation of the Blind, believes there are similar problems with artificial vision.
We’re not interested in technology that promises a lot and delivers little, Gashel says. And an awful lot of technology has been marketed before the promise has been fulfilled. Researchers assume that a little vision is better than using what Gashel refers to as the skills of blindness--walking with a cane, reading braille, or working with voice- recognizing computers. If machines can make people see again someday, he says, they better really do it, not just play around at doing it. Would Gashel, who himself is blind, want to receive the University of Utah researchers’ theoretical artificial eye and read at 150 words a minute, for example? Probably not, and most blind people wouldn’t go for that either, I think--although it’s in the right direction. Still, I’d sure rather read braille at 300 words a minute.
By promoting a neural prosthetic before it’s truly ready, Gashel warns, people may do real harm to the blind. One example he brings up is a device used for schoolchildren with extremely poor eyesight: an expensive video machine that blows up words on a page to many times their original size. Since the children are not completely blind, they are not taught braille. Instead, they are essentially chained to enormous machines that cause them to read far more slowly than either a fully sighted person or a blind person reading braille.
Gashel doesn’t want to pour cold water on research. I just wish the researchers would not use ‘the tragedy of blindness’ to promote the need for their research, because blindness need not be a tragedy. Scientists need to feel they’re helping to resolve some problem that really keeps people down. We’re not pathetic creatures. We’re just human beings.
As someone who is grateful for neural prosthetics, Syrja doesn’t dismiss what their critics have to say. The deaf community has a good point when they express concern about accepting that you’re deaf, she says. I try to take that to heart. If someone takes this speech processor away from me, I’m deaf. But I spoke to a person who is mobility impaired, and I said, ‘What do you think about all this stuff that the spokespeople for the deaf community are saying about how you’re better off just being deaf?’ After all, there are many things I learned being deaf. He said to me, ‘Yeah, I learned a lot of stuff, too, from my handicap, but you know what? If I could walk again, I’d throw it all in a hole.’