If the blind are ever made to see, they might recall with fascination the days when pioneering surgeons inserted electric probes into the eyes of human volunteers and, for an instant, illuminated the dark.
Such experiments, now in progress at the Johns Hopkins Wilmer Eye Institute, may be the rudimentary beginnings of a technology that will enable people blinded by the loss of light-sensing nerve cells to navigate without canes or guide dogs.
Though the achievement may be years away, Wilmer scientists believe they are on a path toward an electronic implant that would complete circuits destroyed by retinitis pigmentosa, a degenerative eye disease that afflicts about 100,000 people in the United States.
For now, in experiments that evoke the eerie excitement of discovery, ophthalmologists Eugene de Juan and Mark Humayun are finding answers to basic questions: Is the damaged retina useless? Can it be electrically stimulated to generate an image? And if so, at what level of detail?
So far, answers drawn from four years of human tests are encouraging.
Blind subjects have seen single circles of light -- like matches ignited in space -- when their retinas were stimulated with tiny probes carrying electric current. It's a hopeful sign that sophisticated technology could allow the diseased retina to perceive at least the outlines of objects.
Someday, these tests may seem quaintly historic, like the first telegraph transmission ("What hath God wrought?") in today's era of digital telecommunications.
Consider the words exchanged recently as de Juan, chief of retinal surgery at Hopkins, held electrodes to the retina of Brian Christensen, a fully conscious man who lost his last vestiges of vision 12 years ago.
"See anything yet?" said de Juan, hunched over a twin-lens microscope that gave him an expansive view of the retina.
"No, it's just a blur," Christensen replied.
"Can you see anything now?"
"Yes. Flash, flash, flash. It was like a white flash. Right over me."
"Like a circle, maybe the size of a quarter."
"You like it?'"
In retinitis pigmentosa, the eye's working parts are intact except for crucial cells in the retina, a membrane in the back of the eye. These cells, called photoreceptors, are responsible for converting light to the electrical code that the brain "sees" as visual images.
Wilmer's researchers envision a chip, millimeters wide, that would sit on the retina and function much like photoreceptors. Early versions would probably work in concert with other electronic components worn outside the body -- say, a button-sized video camera fixed to a glasses frame.
In one version contemplated by the Wilmer team, the camera would capture images and project them to the light-sensing chip. The chip itself would be a grid, subdivided into components that are individually capable of detecting shades of light and dark. Like a plug into a socket, the implant would connect with the surviving neural pathway to the brain.
The idea may prove far-fetched. Neither Wilmer researchers nor the handful of researchers doing similar work elsewhere have any evidence that an implant could supply the brain with enough information to enable a blind person to walk around an apartment safely, much less to read a printed page. But there is a powerful precedent: cochlear implants for the deaf.
Just 20 years after they were introduced, these devices are enabling thousands of people to decipher conversation and, in some cases, to converse on the telephone. They are an encouraging example of electronics bringing sensory perception to people who have lost nerve cells to disease.
"There's an immediate and direct comparison," said de Juan. "The success of that implant was one of the things that really stimulated us."
Asking when the blind will see is much like asking when paraplegics will walk. Despite the almost cocky optimism that pervades the project, nobody wants to make bold predictions about so elusive a goal. But that doesn't stop the team from dreaming about what might happen.
"If I had to guess, I'd say it could be 10 to 15 years before we had something useful," said Robert Greenberg, a medical student and Ph.D. candidate who is the project's electronics expert. He does, however, envision an experimental device within five years -- one that could form the basis for a marketable implant.
Repairing the camera
In many ways, de Juan and his crew are trying to repair nature's camera.
A person's lens focuses light on a two-dimensional film, a concave membrane two-tenths of a millimeter thick and as fragile as wet tissue paper. This is the retina.
Scattered across the retina are 200 million photoreceptors that capture the intensities and hues that make up the scene before one's eyes. Each photoreceptor is responsible for a tiny point of light -- or pixel -- that would look meaningless if viewed by itself. But the miracle of perception depends also on the brain, which blends the pixels into the seamless picture that constitutes a person's visual reality.
Retinitis pigmentosa destroys the photoreceptors in a relentless assault that often begins in childhood and causes blindness by middle age. The destruction starts at the edges and moves toward the center, causing an ever-narrowing field of vision.
Someday, scientists may figure out a way to restore sight by triggering the regrowth of nerve cells or by transplanting new ones onto the person's retina.
Humayun was a young medical student at Duke University when he started thinking about a different fix -- a technological one. At the time, he was working in the laboratory of a first-year faculty member who was building a reputation as a sure-handed eye surgeon with an innovative bent.
Humayun, a native of Pakistan whose family moved to Potomac when he was a child, told de Juan that he had grown weary of the cell-biology project that was occupying him. Offering a bold alternative, he suggested a long-term study that could lead to an electronic eye.
The idea wasn't entirely new. Humayun had read of efforts at several centers, including the National Institutes of Health, to develop technology that would connect an external optical device directly to the brain's visual cortex, bypassing the eye entirely.
Theoretically, a direct connection made sense for people who had lost their optic nerve to glaucoma. In cases like these, stimulating the retina was useless because disease had severed the electrical pathway between eye and brain.
But that connection remains intact in cases of retinitis pigmentosa. Stimulating the retina takes advantage of the marvelous wiring that nature installs -- wiring that distributes information to receptors scattered across the hills and valleys of the visual cortex.
Rather than dismiss the idea as crazy, de Juan saw exciting possibilities.
A partnership was born.
Humayun completed his medical residency, earned a doctorate in bioengineering and joined de Juan in animal experiments that led to the human tests now in progress. In the early experiments at Duke, they found that they could trigger a response in a rabbit's visual cortex by applying low-powered current to the retina.
On an electronic monitor -- similar to the ones that measure brain activity in humans -- they saw responses that might well have corresponded to bursts of light across the rabbit's consciousness. The rabbits, of course, couldn't describe what they saw, so the next step was to solicit human volunteers.
The first was Harold Churchie, now 69, a Sharpsburg man who had spent three decades running the snack bar at the Hagerstown courthouse. Like his twin brother, Carroll, he was destined to go blind because he had inherited the genes for retinitis pigmentosa.
Churchie wasn't seriously afflicted until he was in his 30s. That's when he started to lose sight of everything in the corners of his vision. The world continued to close in, and by his early 50s, he was left with a small circle of central vision.
"If I rode down the road with you, I could see a car coming the same as you. But I couldn't see that motorcycle on the side," he said.
In 1992, a few years after his sight had disappeared, his doctor at the Wilmer Eye Institute told him about de Juan's and Humayun's work at Duke, and about their need for volunteers.
"I said, 'Well sir, it might not help me, but if it can help some child somewhere down the road, I'm all for it,' " Churchie recalled.
With a local anesthetic applied to one eye and a metal ring holding his lids wide open, Churchie lay in an operating room at Duke as de Juan pricked a hole in his eye and inserted a fine electronic probe.
"When they put the electrode in and moved it around, I could see it moving around. From one corner to the other," he said.
Said Humayun, "You couldn't ask for anything better. He could tell us exactly where we were stimulating. When we moved a little bit, just half a millimeter, he could say, 'You're moving!' "
The experiment broke ground. It showed that a current applied to, say, the right-hand corner of the retina triggered a flash in exactly that part of a person's vision. This provided hope that an implant, supplying many points of light at once, could map an image that would correspond to an object's place in space.
By 1994, Hopkins had lured the two surgeons from Duke. Their research into electronic vision was just one attraction.
De Juan, for instance, was one of the only surgeons in the world who performed risky surgery on premature babies whose eyes were damaged by excessive oxygen in the incubator. He had also pioneered "submacula surgery," the removal of scar tissue from a delicate and hard-to-reach area behind the retina.
Humayun, trained in many of the same areas of retinal surgery, brought his expertise in biomedical engineering -- the integration of technology with living systems.
Since coming to Hopkins, the surgeons have experimented again on Churchie along with nearly a dozen other people. One key question is what happens when the retina is stimulated in more than one place simultaneously. Do the images blur as one, or become distinct points of light in a person's field of vision?
In a recent experiment, they challenged Brian Christensen with a thin probe that was actually three wires bundled together. Christensen, from Watertown, Conn., grew up with poor night vision, but his daytime sight didn't erode until he entered his 20s. Then, it went fast.
When he enrolled in the experiment, he knew that he might not enjoy the fruits for many years -- maybe never. "Who knows, maybe someday I'll be able to walk by myself with some vision," said Christensen, 34, who uses a cane and assembles knapsacks in a workshop for the blind.
In a Hopkins operating room, nurses secured Christensen to a gurney with a tight, cloth wrap.
At right angles to one another, de Juan and Humayun looked through microscopes into the interior of Christensen's left eye. A healthy retina is pink, but Christensen's was a ruin of black grains.
Robert Greenberg, the student pursuing degrees in medicine and bioengineering, manipulated a computer panel that enabled him to activate up to three electrodes at once. He also varied the current's strength, hoping to discover how much energy was needed to generate an image.
His hands almost inhumanly steady, de Juan held the electrical probe to the patient's retina and slowly moved it from place to place.
"I saw a flash," Christensen said early in the experiment. "I could tell where it was, right over me."
At this point, he was seeing a single flash that corresponded to the firing of a single electrode. When Greenberg turned up the current, the light got brighter. And when he activated two electrodes, Christensen saw them as one until his eyes adjusted.
"It's two flashes," he said.
"How far apart?"
"Maybe a fingertip."
Clearly enjoying himself, de Juan looked at his patient and said, "You talk about people who really go out of their way. Nothing is more unselfish than this kind of thing."
On another day, Churchie returned for a second test. This time, the surgeons challenged him with an electrode that was actually a five-by-five array of stimulation points -- 25 pixels in all.
From his control panel, Greenberg activated a combination that spelled out the letter "H." Churchie saw a "U" -- a perception, perhaps, of the top half of the letter "H." Later, Greenberg theorized that Churchie saw the letter incompletely because his eye was jiggling slightly.
"Our focus now is to have an implant that we can actually fix to the eye to avoid that problem," Greenberg said.
Brain fills in blanks
Fashioning a visual prosthetic that works as well as hearing implants will require more than technical wizardry. It will also require a better understanding of the brain's flexibility -- whether, for instance, it can process a crude assortment of visual cues and fill in enough blanks to form the image of, say, a door or dog.
To some extent, the brain already does this in the sighted person. "When you have something that moves very fast in your field of vision, it looks much sharper than it actually should," said Dr. Gislin Dagnelie, a member of the Hopkins team who studies the physics and psychology of visual perception. By all rights, a moving car ought to appear as a smear.
"The brain is actually doing a lot of reconstruction because we know what an object actually looks like," he said.
The most sophisticated implant will have to depend on the brain to perform more impressive acts of reconstruction.
"We really don't know what people will see," said Dagnelie. "We hTC don't know if they will see enough to read with a magnifying glass or watch television.
"We really don't know."
An electronic eye
An electronic implant might one day restore sight to people afflicted with retinitis pigmentosa, which destroys photoreceptor cells but spares the rest of the neural pathway to the brain. At the Johns Hopkins Wilmer Eye Institute, scientists are conducting tests they hope will lead to such an implant. They have not yet worked out a specific design, but here is one possibility.
How a person would see an "E."
1. A button-sized video camera mounted on a glasses frame would capture the scene before one's eyes -- in this case, the letter "E."
2. A laser mounted on the camera would project the image onto a light-sensitive, electronic implant secured to the back of the eye by tiny tacks. Laser would also power the implant.
3. The implant would be a grid of smaller squares that are individually sensitive to light. Each square highlighted in white above would capture a small piece of the "E" and apply electric current to a corresponding electrode.
4. Electrodes would transmit the pulses to neighboring ganglion cells, neurons that relay the information to the optic nerve and, ultimately, to the brain. The brain perceives the input as the letter "E."
Pub Date: 9/15/96