From Nature, Machines
Ralph Etienne-Cummings studies animals in order to make devices that could one day help paraplegics walk, the armless feel and grasp, and machines see and fly like insects.
By Michael Anft
Here's another one: More than 250,000 people are paralyzed in the United States, most from the waist down. You want to create a device that can get them moving again, one that doesn't repair the spinal cord but instead works with undamaged nerves in the spine that, if tapped correctly, can generate the precise patterns of muscle movement used in walking. You need your device to be energy efficient, implantable, and tiny. How do you build a portion of a spinal cord in a microchip? Might the swimming patterns of an eel yield some clues?
Infusing machines with the intricacies of biological systems developed over millions of years of evolution is enough to vex even the most confident engineer. Just as astrophysicists have spent a mere blip of the age of the Earth divining that the universe is about 14 billion years old, engineers who mine biology to develop new systems have a lot of chronological turf to cover in a relatively short bit of time. To make their ideas concrete, they often must collaborate with biologists who better understand the ins and outs of how animals work. And they must downplay the expectations of a culture weaned on sci-fi technophilia — visions of cyborgs with flame-shooting prosthetic arms, or human heads outfitted with digital brains — to answer more basic questions about biology and functionality. If you were able to line all that up and get to work, you might be able to achieve a series of breakthroughs: machines that offer all the benefits of nature.
At least that's what Ralph Etienne-Cummings is banking on. A professor of electrical and computer engineering at Johns Hopkins, Etienne-Cummings tackles problems that marry the biological wisdom of the ages with a futuristic, high-tech derring-do. Propelled by an engineer's burning desire for solutions, a scientist's curiosity, and the spirit of a kid who grew up on sci-fi films, he negotiates the shotgun wedding of two worlds that are poles apart.
"The interesting question for me is, how do you take something that is completely different from biology and interface it with a biological system?" says Etienne-Cummings, a compact man who delivers speech swathed in an exotic island accent. "For me, it's all about understanding where electronics and biology meet. How can you use an aspect of one to control the other, so devices can either work within biological systems, or work separately, but still like those systems?"
A full professor at the age of 42, Etienne-Cummings enjoys a reputation internationally as a leading researcher in biomorphics, the engineering field that uses biologically inspired principles to make prosthetic limbs, robots, and other devices. His niche is neuromorphics, the study of how animals' muscle control and sensory systems work so engineers can model them in silicon — or, put another way, jam-pack integrated circuits into microchips that can perform the same functions as nerve cells.
Even though Etienne-Cummings works in a high-tech bubble, surrounded by slabs of circuit boards and rows of video screens, mimicking animals is old school. Humans have been studying animal behavior since the days of cave dwellers, when anticipating or copying the movements of the hunted led to dinner more often than not. More recently, the Wright brothers spent months observing birds in flight to see how they maneuvered their wings during turns.
For all of the intelligence and complexity of another modern man-made marvel — the digital computer — it is no match for nature. Animals move and react in elegant ways that machines can't yet mimic. Their systems may "spike" into levels that can be measured and replicated digitally, as the nervous system does when excited, but more often they behave automatically or instinctively while expending a low, analog hum of energy. Replicating the body's functions can mean being creative with lower-tech solutions as much as putting together endless series of Os and 1s. And yet those "simple" biological systems behind the body's movements are efficient enough to give biomorphic engineers hope they can be used as models for devices that guide robots or restore function in humans.
Humbled though he is by the intricacies of nature,
Etienne-Cummings is among the hopeful. Part of his success
in moving emerging biomorphic systems forward is his choice
of world-class biologists as collaborators. "He works in a
funny field, to tell you the truth," says Steven Hsiao, a
professor of neuroscience at Johns Hopkins' Mind/Brain Institute.
"Guys like him don't really experiment on their own. It's
hard to be great at both the bio side and engineering. So,
the weird thing is you need people in other disciplines to
help you do your work." Etienne-Cummings combines his
talents with those of Hsiao and others at the Mind/Brain
Institute to understand how the nervous system converts
stimuli to sensations, and how neural information travels
from the spine to the brain and back. Along with others at
the Applied Physics
Laboratory and the Whiting School of
Engineering, they are part of a multiuniversity effort
to develop an advanced prosthetic arm that can feel, grasp,
and monitor itself in space.
Ralph Etienne-Cummings is a leader in biomorphics, a field
that marries biological principles and engineering. His
current quest is to give paraplegics feeling and movement
in their legs.
Photo by Will Kirk
But what gets Etienne-Cummings the most attention from engineers and federal funding agencies is his decade-long quest for a spinal implant that will help people paralyzed from the waist down regain movement and sensation. By taking on one of those problems, he's attempting to answer a long-standing and high-profile bioengineering question: How do you reanimate the human form?
When a person's spinal cord is severed, signals from the brain can't reach regions below the break. People who have a lower-spine injury also lose the ability of nerves in their legs to "talk" to each other, making the coordination necessary for walking impossible. Etienne-Cummings has sought to replicate the nerves that fire when a person walks — bundles of so-called controller nerves that create central pattern generators, or CPGs.
In a healthy person, the generators feed signals into nerve cells in the spine — cells that remain even after a spine injury — and then respond to the output. The nerve cells learn to balance and coordinate the inputs and outputs in the proper sequence for locomotion. Etienne-Cummings challenged himself to come up with a way to tap this natural process to get paralyzed people moving again.
First, he studied how a machine modeled after a walking human works. He joined forces with Anthony Lewis, a roboticist friend, and Avis Cohen, a biologist and neuroscientist at the University of Maryland, to create artificial legs that could be commanded via microchips that mimic CPGs to walk seamlessly on a treadmill. But there were still several engineering hurdles to climb. "We showed we could make a robot walk, but even if I could build a model of the spinal cord, could I stimulate the muscles in the right way and in the right order to get a person to walk? That's what we had to figure out," Etienne-Cummings says.
He and Cohen began studying the spinal cords of lamprey eels. Besides having a superhero-like power to regenerate damaged spines, lampreys have 100 or so CPGs that govern how they move when they swim. Etienne-Cummings and Cohen wanted to create a device that would mimic the CPGs and persuade existing nerve centers in paralyzed people to send messages that would allow them to walk again.
Aided by grad student Jacob Vogelstein, Eng '07 (now an assistant professor of electrical and computer engineering), they started filling in pieces of the puzzle. First, they analyzed the electrical output of the spine of a swimming lamprey. Next, they used a microchip device to control how an eel swims, which enabled them to measure electrical output along the eel's spinal cord. Finally, they started work on a more refined chip that would emulate those nerve signals. "The lamprey helped us learn to close the loop of movement, of getting, say, the left side of the spinal cord working just so to get the right side to respond and move," says Etienne-Cummings. "If we can do that, we thought, could we do it in a more complex animal?"
The eels imparted one more lesson for researchers. Though the animal is able to regenerate spinal cord cells, the eel never moves the way it used to once its severed spine grows back together. "They develop some weird swimming patterns. They might swim so haphazardly that they are functionally useless," Etienne-Cummings says. "So, we have this animal that is built to regenerate, but it doesn't do it exactly right. What does that tell us? Well, what we need to figure out now is how electrical stimulation might work to aid in that regeneration, so we can get the animal moving." Researchers at institutions across the country, including the Johns Hopkins-affiliated Kennedy Krieger Institute, are experimenting with techniques that involve spinal stem cells to restore movement. It's possible that the stem cells might have the same imprecise effects in people as regenerating nerve cells do in lampreys. "Which would mean that there is still a role for certain types of electrical stimulation to guide the stem cells' growth," Etienne-Cummings hypothesizes. "We're likely going to need a combination of techniques in tandem to make the spine work again."
The tack he and others have taken is a departure from other experimentalists. Scientists at Case Western Reserve University, in Cleveland, have implanted nerve-like wires in paraplegics, who control the wires using a computerized box attached to their belts. This "functional electro-stimulation" model can get people walking by pushing buttons to lift or lower their legs. While hailed as a step forward for science and for patients, functional electro-stimulation also takes a step back: Patients tire quickly because the method moves muscle fibers in the opposite order than nature does. In essence, the machine works against the body as much as it does with it, tiring muscles that have been programmed by nature to respond differently. Another drawback: Patients must undergo extensive surgery to implant the wires.
The technique Etienne-Cummings and others are developing, called intro-spinal stimulation, will fire nerves to move muscle fibers in the order nature intended, and it won't require wire implants or an external controller. A better idea, but one that still needs some bugs worked out. Walking requires constant, simultaneous computations, coordinated movements, and feedback to the brain. Etienne-Cummings and crew must invent a CPG that does all that.
"Here's the thing, right? You might be able to help someone turn right as they walk, but unless their legs know that they're turning right, and unless the brain gets that message, then they don't know how they're doing. They won't have much confidence. So what you have to do is create this continuous communication between the CPG in the spine and the brain," Etienne-Cummings says.
Although the mystery of how to "close the loop" enough to aid humans remains, Etienne-Cummings says there is less of a mystery between biology and electronic circuitry than one would think. Electricity is part of nature, too, he points out. "If you look at the properties of how transistors work — how they move charge from one place to another — the physics of that is similar in the way ions move back and forth between nerve cells," he says. "I can put enough of these transistors together to start replicating the same kind of electrical functionality we see in living cells."
Etienne-Cummings is a product of nature himself. While growing up in a remote island chain where television didn't arrive until he was nearly 10, he spent days holding strings connected to hooks in hopes they'd get connected to fish. "We never used rods, just lines," he says. "We'd throw the line out there and keep a part of it in the palm of our hands until you'd feel the little tug. Then, we'd set the hook. It was pretty primitive."
A sun-swept string of granite-and-coral dots in the Indian Ocean, Seychelles is a former British colony just east of Africa that runs on tourism and tuna fishing. Like most of its ethnically mixed population of 80,000, Etienne-Cummings grew up on the steep hills that mark the island of Mahé, living in a home as large as a good-sized room. The son of teenagers, young Ralph was raised mostly by his grandmother, while his mother lived in England, caring for other people's children and sending a big chunk of her paycheck back to the island.
Much of Ralph's time was spent playing cowboys and Indians in the bush along with his many cousins and friends, who learned about the battle for the American frontier while watching spaghetti Westerns at the local cinema, the Odeon. Ralph also nurtured a passion for science fiction, watching movies like Star Wars and 2001: A Space Odyssey, even as he and his friends enjoyed the uncomplicated pursuits of playing soccer and climbing trees. Etienne-Cummings does little to dispel the image of island life as anything less than idyllic: "We just ran around a lot, you know?"
Even though the island was hardly flush with currency, it did feature private schools that outperformed the national ones. "You had to register as soon as your child was born," recalls Marguerita Etienne-Cummings, Ralph's mother. "I pushed for Ralph to get into one. I always believed in a strong education." She paid for it out of her U.K. earnings, as she did daily sessions with a tutor for Ralph — sessions he frequently ditched.
"He was the type of kid who would hide and then jump out and scare the wits out of you," his mother, who now lives in Wilmington, Delaware, adds. Despite "having some of the devil in him," she says, he excelled at school, getting As starting in first grade. He showed a curiosity for learning how things work, breaking toys open to see what made them run. "It was infuriating," his mother says.
Although he would have trouble in formal English and French classes, the Creole- and English-speaking Ralph was a math and science whiz. Early on, there were signs he might end up building things, instead of breaking them. He could solve practical, real-life problems, like when the intermittent reception of the family's short-wave radio threatened to make static mush out of the broadcast of a game featuring his beloved Liverpool soccer squad. Ralph and a cousin climbed trees, putting wires and umbrellas in them to extend the radio's tiny antenna. "If you walked in my neighborhood, you'd probably look up into the trees and say, 'What the . . . ?'" he says. "It seemed to help, at least a little. Maybe the signal didn't fade out as often."
When he was 8, his mother came back to Seychelles with a U.S. Air Force man whom he would soon call "Father." Although Ralph maintains ties with his biological father, an optician who lives in Seychelles, Herman Cummings became the man of the house after marrying Marguerita and moving Ralph and his younger brother, Kevin, into a middle-class island home.
In 1977, after the islands had undergone a revolution, the new government banned private schools, forcing a major change in Ralph's life. "They wanted all schools to be equal, but instead of making the worst schools like the best, they did the opposite," says Ralph. More troubling for his mother was that kids in the ninth grade were sent away from their parents to a boarding school on the island. "The people who had enough money sent their kids to England," she says. When Ralph reached age 12, he became one of them. Despite the happy childhood, he didn't exactly weep at the prospect of leaving home. "To me, it was like, woo-hoo!" he recalls. "I loved home, but cabin fever is real on an island."
With his parents in Seychelles, where his father helped manage a U.S. satellite tracking station, Ralph was on his own to study at a strict Benedictine school in Kent. It took him a couple of years to catch up with the other kids in the classroom, but by year three, he'd passed most of them. "The good thing about it was there were people at the school from everywhere — Nigeria, Hong Kong, America. It was diverse, like Seychelles," Etienne-Cummings says. Eventually, he became comfortable in England, playing field hockey, rugby, and soccer, and gaining a reputation as a serious runner. Toward the end of his high school years and in college, Etienne-Cummings' best times would have qualified him for the Seychelles team in the 400 meters at the 1988 Olympic Games in Seoul, South Korea. But Seychelles' ties to communist North Korea led the country to boycott. "My letters to the sports ministry informing them of my times were never acknowledged," he says. "No hard feelings."
Etienne-Cummings' parents moved to New Orleans the year before he graduated from high school. He followed them, but not directly, making a detour to see an uncle in Cincinnati. He aimed to walk in his father's footsteps and become an Air Force man. "I wasn't trying to get into university," he says. "I took the test to get into the Air Force, and did well. But because I wasn't an American citizen at the time, I couldn't get the jobs that I wanted, like in electronics or working with aircraft. They wanted to make me a cook or something."
A cousin had entered a NASA-backed program for aspiring African-American scientists and engineers at Lincoln University, a historically black school near Philadelphia. Etienne-Cummings applied to the school, which, four years later, would graduate 25 percent of all black physics undergrads in the United States, including him. "Lincoln was where I got the idea that, hey, I'm different," he says. "It was all black. I'd never seen all black before."
He majored in physics, but he also learned about electronics and how to use chemicals to etch the pathways in silicon that are essential to making the circuits behind microchip technology. Etienne-Cummings proved to be a quick study. "I gave him a pile of physics papers, and he returned two days later and said he'd read them all. 'Sure you did,' I thought," says Irvin Heard, Etienne-Cummings' mentor at Lincoln and now a lecturer in physics at John Jay College of Criminal Justice, in New York. "But as I quizzed him, it became apparent he had read them and understood them thoroughly. Every time I asked him a question about materials science, he was the only one in my classes who could give me the answers." His intelligence was put to work immediately on uncovering the esoterica behind electronics and circuitry. He learned how electrical properties in certain chemicals could make circuits do certain things, including fire off at certain rates.
That interest in chip technology led Etienne-Cummings to seek a PhD in electrical engineering at the University of Pennsylvania. "At Penn, people were less interested in how you could manipulate silicon as they were in how you could apply it," he says, adding that creating chips that could mimic biology was among those applications. "It was a natural progression for me." Natural, but because he had majored in physics, not necessarily a breeze. He had to double up his course load and take undergrad engineering courses to learn about the mechanics of circuitry.
During a sophomore class on designing chips from transistor-based integrated circuits, a group of his fellow students sought him out. They were having trouble squeezing their design into a transistor space and thought they might persuade the grad student whose test grades broke the curve to help them. The mission involved swallowing some pride, so they drew straws. A 20-year-old woman named Shamita ended up with the short one. She and Ralph would marry a few years later, but she remembers the moment she asked him for advice as a bit scary. "Ralph's a very nice guy, but he knows everything, which can be intimidating," she says.
After learning some of the basics, Etienne-Cummings took the chance Penn afforded him to get out in front of the integrated circuits-based electronics boom of the late 1980s. Only a handful of professors nationwide (including Andreas Andreou, an electrical and computer engineering professor at Johns Hopkins) taught about integrated circuits. One was Jan van der Spiegel at Penn, who applied the emerging technology to systems inspired by biology. Etienne-Cummings had found his calling.
"Ralph learned all about electrical engineering, of course, but he was really fascinated by the biologics," says van der Spiegel. "He had a lot of intellectual curiosity and wanted to explore other areas. You could see it in his dissertation, which was developing a smart retina that could capture images in motion and process them, just as an animal's does. It was one of the first chips that could do that."
In the van der Spiegel lab, he and others built a circuit model of the brain — an analog neural computer, complete with neurons, synapses, and dendrites, the mechanical agents behind sensation and movement. Etienne-Cummings contributed by building circuits that mimicked human neurons that performed processes of the brain.
After graduating from Penn in 1994, Etienne-Cummings took an assistant professor's post at Southern Illinois University and inaugurated an ongoing annual seminar he gives through the Institute of Neuromorphic Engineering, the group he formed and chairs in Telluride, Colorado. While on a hike in the Rockies, Andreou and another Hopkins professor who has since left recruited Etienne-Cummings to come east.
Etienne-Cummings came to work at Hopkins in 1998 and immediately got moving on a vision system that would cause a fast-moving toy robot on a track to avoid obstacles on its own. The idea was to make the robot "see" and develop what Etienne-Cummings calls "situational awareness" and respond to it. Before long, he was applying the neural computer concept he helped develop at Penn to a new idea: producing neural networks that create signals for controlling locomotion, a concept that led to the robot/treadmill successes. The development gave him hope that a single, analog silicon chip four millimeters square and using around 10 microwatts of power — much smaller and much more effective than digital converters that had been considered for such use — could tap into the spine's residual nerves and help people with damaged spinal cords regain function.
Along the way, he has picked up projects that study how bats process hearing, how the brain deals with touch sensations, and how houseflies convert visual information into hazard-free flight (to aid in the development of unmanned aerial vehicles), then created silicon chips to perform the same functions in several machines. In the works is a project to develop ultrasound technologies that could be used in more types of medical imaging and, possibly, to burn out tumors.
During his inaugural lecture after being named a full professor last November, his mother and five other relatives and friends came from Seychelles to see him speak. So did van der Spiegel: "At Telluride, he usually focuses on one thing or another. But at this lecture, he really gave a comprehensive overview of what is possible. He's taken the field much further, particularly in regard to neuromorphic vision and robotics. It's amazing what he's been able to accomplish."
The attic office/lab that Etienne-Cummings runs in Barton Hall is all angles, video screens, wires, and circuit boards. The slants come from the Gothic shape of the building's roof, which gives the long lab room a bit of a mad-scientist tilt. A quick visit with Etienne-Cummings' grad students doesn't help shake that feeling. There's no cyborg assembly line here, no high-tech Frankenstein in the offing, but the students — mostly PhD candidates — aren't merely playing with toys that further their education, either.
Like their mentor, Etienne-Cummings' protégés often seek clues in nature. And again like him, many students come from Africa, a result of both his active recruitment of talent during trips there and his ability to entice students from around the world who have heard of him.
One student in electrical engineering from Nigeria hooks up a circuit board that models the retina of an octopus — a primitive system that students use to learn how to replicate a function of the body in silicon. Another Nigerian PhD candidate in electrical engineering continues to work on his invention, a device that measures electrical signals in the body and combines the functions of electrocardiograms, encephalograms, and other medical tests into one box of circuits and wires. Another grad student works on a sensory substitution project that will develop a glove blind people can wear that "reads" colors and then feeds "vibrations" back to provide visual information that can help them decipher objects.
Downstairs in the New Engineering Building, Alex Russell works toward his PhD in electrical engineering by helping Etienne-Cummings and other researchers develop a prosthetic limb that moves and feels like a real arm. Russell's role: to concoct algorithms that will measure the signal the body sends to the brain when a person touches something. He met Etienne-Cummings while a student at the University of Capetown, in South Africa, where Etienne-Cummings was teaching on a Fulbright scholarship. After earning his bachelor's degree there, Russell followed Etienne-Cummings' trail to Johns Hopkins. "He offered me something I have a lot of interest in but I couldn't get in Capetown — a chance to work in engineering, but while working on biologically derived solutions to problems," Russell says. "He's quite a draw."
The work of the Etienne-Cummings lab may be far flung, but for him, the spinal-implant project is the most intriguing one. He and a Canadian researcher whom Vogelstein met during a bioengineering conference have already shown that such a device will move the legs of a cat whose lower spinal functions have been temporarily knocked out by drugs — a result that has caused a stir of its own among researchers around the world. But there are a lot of problems to work out before the problem is solved to the point where humans can benefit.
"It's a moving target," Etienne-Cummings says. "We have a good idea right now of what we can do. We know we can stimulate nerves, move the legs in the right way. Now, we have to close that loop between biology and electronics."
That might happen soon. It's not out of the question that a device could be ready for human use within 10 years, Etienne-Cummings adds. He and Canadian Vivian Mushahwar, a professor of biology at the University of Alberta, hope to test the implant on spinal tumor patients within the next five years, then perhaps begin clinical trials. The advantages the device could have for people with no nerve activity below the waist drive him, he says. He envisions a time when the silicon implant, combined with electro-stimulation, stem cell therapies, and other treatments, reanimate limbs so they move very much in the way Mother Nature intended.
"It's not just about walking," he says. "That's the sexy thing people talk about — maybe it's part of our Judeo-Christian heritage, this need to heal in a big way. But what about going to the bathroom, or feeling sensations below the waist? Walking is not as high up on the list of things paraplegics would like to have back as you'd believe."
While the spinal implant motivates him to work hard, he's on no crusade. One of the interesting things about him, colleagues and students say, is that he is rarely rattled or driven to distraction, and that he's open to doing a lot of different things. Shortly after he entered the electrical and computer engineering PhD program, Vogelstein asked Etienne-Cummings to be the mentor for his thesis project. The two became friends and colleagues, working on the implant project together with Cohen. Now they socialize — mountain biking, restaurant hopping, snowboarding, skiing, or barbequing at Etienne-Cummings' Washington home. (Such events often feature fresh snapper flown in from Seychelles.) Vogelstein says that Etienne-Cummings evokes a rare brand of smoothness, whether he's calmly explaining a complex concept, or driving his white Mercedes roadster — a laidback figure behind the wheel, under a black porkpie hat.
Etienne-Cummings keeps at least one night free each week for Shamita, a patent attorney. They often see films of all kinds, and have a special soft spot for independent movies, but sci-fi is still one of his favorites. And though he is immersed in the tough, nuts-and-bolts work of making the unlikely happen, he isn't averse to injecting a little fantasy into his presentations now and again.
"The imagination behind science-fiction movies or stories can lead to a lot of things that we do in the lab," he says. "So I always finish my talks with the piece of video in my seminar from [the] Star Wars [trilogy], the one where Luke gets a new arm. It provides some inspiration for applying the things that we're working on."
Michael Anft is senior writer at Johns Hopkins Magazine.
The Johns Hopkins Magazine |
901 S. Bond St. | Suite 540 |
Baltimore, MD 21231
Phone 443-287-9900 | Fax 443-287-9898 | E-mail firstname.lastname@example.org