Mindful Motion: Miguel Nicolelis and Mind-Powered Robots; and Creating Science Cities in Brazil and Beyond
Steve: Welcome to Science Talk, the weekly podcast of Scientific American for the seven days starting January 16th, 2008. I'm Steve Mirsky. Last week, the mind of a monkey in North Carolina controlled a walking robot in Japan. The mind that created interfaces between intelligent beings and robotic limbs belongs to Miguel Nicolelis. Scientific American editor Christine Soares recently sat down for a conversation with Nicolelis in his office at Duke University, where he is the codirector of the Center for Neuroengineering. They talked about the organic, robotic neural interface research and its implications for prosthetics; as well as some exciting plans for a grand sociological experiment in Brazil, the creation of an entire science city.
Soares: In 2004 you gave [a] really impressive demonstration of how your work could lead to prosthetics controlled by the brain. Your star monkey, Aurora, learned to play a simple videogame with a joystick, but you were also essentially downloading the neural signals from her brain to control the robot arm.
Nicolelis: Yeah! In that sense what we were doing was to record the brain activity that Aurora was producing to generate arm movements and after a little bit of training, both Aurora and I, we were able to basically get these signals to be decoded in real time and translated into digital commands that could be used by a robotic arm to generate movements that Aurora was imagining. So at a certain point, Aurora realized that she didn't need to move anymore, she could just imagine the movements and this interface that we created, this brain-machine interface that we created was able to enact her will and generate the movements that she needed to produce to win the video game.
Soares: So this could really be an example of how the language of the brain can be translated to move a prosthetic for a person who is a paralyzed?
Nicolelis: Yeah, that was one of the ideas. First we were interested in understanding how this language of the brain is produced and what is the code and that's what we were doing at that time; and we learnt that by using the simple algorithms we could read these signals and control a mechanical device, so there opened the possibility in the future for patients who are paralyzed to use their brain activity to directly control a variety of devices.
Soares: So it's some very fundamental neuroscience questions that you are answering, but also some very practical applications?
Nicolelis: Yes. The idea was to design a prototype that could handle both issues, how do you address fundamental questions about how brain signals operate and how can you actually use this knowledge and this information in technology developed to address this issue to generate some practical application.
Soares: And how've you been following up on the Aurora performance?
Nicolelis: Yeah! Well, we have now used the same idea to study other types of behaviors. For instance, locomotion and we've learned that we can do the same. We can read signals from motor and sensory areas in the brain that are involved in the generation of the motor program to walk, and we're able to read
this [these] signals, decode them and send them to a device—a robot, a bipedal robot that actually starts walking like the monkey. And we are planning to do now a series of experiments that will demonstrate the power of this interface by getting the monkey not only to control in real time this robot—that is not going to be here in the United States, it's going to be in Japan, in Kyoto, the ATR Robotics Lab—but also get feedback signals from the robot back to the monkey, to see how the monkey interpret[s] the fact that she is now or he is now controlling a device about 10,000 miles away in real time.
Soares: That's pretty impressive. I think with Aurora, the robot arm was in the next room.
The Aurora's arm was in the next room. We had done experiments before Aurora with [a] couple of other monkeys in which one robotic device was at M.I.T., while the monkeys were in Durham, so we learned a little bit about how to do that. The challenge here we discovered in Kyoto is much larger because we are talking about a lot more information being transmitted; and is [it's] different from the experiments we did in 2000 with M.I.T. This is now going to be a bidirectional connection. So, it will be really a blend of the monkey's brain with this robotic device.
Soares: Great—you are going to be feeding sensory information back to the monkey.
Nicolelis: Yes. We are going to get sensors that are located in [the] feet of this robot to send feedback information back to the monkey, and we are going to give the monkey an opportunity to experience what it is to control a device that is in a different continent.
Soares: That's incredible! We look forward to seeing those results and that video (laughs).
Nicolelis: Sure; it's going to be fun.
Soares: You're also working on counteracting the symptoms of Parkinson's disease, or at least understanding them.
Nicolelis: Yes, we—in collaboration with the lab of a good friend of ours, Marc Caron here at Duke—we have actually looked into the brain of transgenic animals that Marc created that can do, you know,
to certain manipulations that it can [be] induce[d] by first genetic and then pharmacological [means]; you can actually produce symptoms that are similar—some of them are similar—to what you're seeing in Parkinsonian patients, and you can treat these mice and actually get these mice everyday to express these symptoms and recover. By recording the brains, you know, since they are normal until the moment in which they become Parkinsonian, we actually learned what happens in that transition; and since the transition happens in 45 minutes, you know, everyday, we actually accumulate a lot of information [on] and how a normal brain evolves to become Parkinsonian or at least Parkinsonian-like. And now we realize that there are ways to measure this brain activity, for instance, in a motor cortex and find that these cells are becoming more synchronous, they are fighting more together than they used to be, almost like a mild seizure, a mild epileptic seizure; it is not a seizure, but the synchronous fighting resembles that. So we started testing a lot of ideas now on how to treat that, and we are just finding that we may have an opportunity to desynchronize these neurons and get these animals to improve. And that's what we're studying right now, we're trying to come up with methods that to some degree link this research to the prosthetic research to desynchronize this [these] neurons and get the mice, you know, to be able to move and to get rid of this [these] Parkinson symptoms by doing so.
Soares: So, is there a common thread in all the different things you're working on?
Nicolelis: Yeah, I think the common thread of all these stories is the attempt to understand how the brain works by looking at the population level and the neuronal circuit level. I mean, of course, neuroscientists have been devoting almost a century of work to understand how single neurons operate, the physiology of single cells; but what we realized lately in the last couple of decades is that to obtain behaviors, to generate behaviors, the brain basically needs to recruit populations of neurons,
this [that] stood across many structures that come together for a moment in time to generate an output. And actually the game seems to be there, you know, so what we have been doing in all these lines of research is to focus on the operation, the physiological operation of circuits, not single neurons, and try to understand what the emergent properties [are] that come from a combination of hundred of cells at once.
Soares: The whole system!
Nicolelis: The whole system. Yeah, instead of just looking piecemeal one at a time or each structure of the brain at a time, we want to basically measure and quantify the operation of the whole brain or most of it, as it happens, you know,
that (unclear 8:32) in the case of the motor control, in those 300 to 500 milliseconds that precede the onset of movement. And we really wanted to see how the circuit comes together and generates the signals that are required for movement to emerge.
Soares: And you're not just doing this at Duke; you actually have several labs around the world, right?
Nicolelis: Yeah! Right now, we have like a network of laboratories that we have established to work on these projects. We have laboratory at the École Polytechnique Fédérale de Lausanne—E.P.F.L.—in Switzerland that is doing some work like that in locomotion in rodents. And we have two major laboratories in Brazil, one in São Paulo and Sírio-Libanês Hospital that is participating in a collaboration with us; and, of course, two laboratories in Natal in
the northeastern Brazil where we are creating an institute, an international neuroscience institute.
Soares: Great! And that's happened pretty quickly just in the past few years?
Nicolelis: Yeah, basically this idea of creating this global network of collaborators and labs is basically unfolding in the last three to four years, particularly
to[the] Brazil effort in the last four years.
Soares: Yes. You just inaugurated earlier this year a school and a health clinic and the institute itself, the research labs.
Nicolelis: Yeah! So, because the institute in Brazil—the Edmond and Lily Safra International Institute—is basically, it's not only about neuroscience and not only about doing research like we do here at Duke or in other institutes around the world; it's about using science as an agent to social transformation of a community. And so in addition to creating a neuroscience institute that is linked to the best neuroscience institutes in the world, we created the first science education school in Brazil for children;
that [which] is now enrolling 800 kids from the worst school district in Brazil, or one of the worst, and hopefully by the end of the year we'll have around thousand kids participating[in] this after-school science education program. We just completed the construction of [a] women's and child[ren's] clinic that is next to that school to basically also act on translating all sorts of ideas in medicine and in neurology to healthcare services to the community. And now we are starting the construction of what we would like to call "the campus of the brain" that we consolidate all these activities and create the whole mission of the institute.
Soares: Talk about the area where you're building this. It's a very poor region in northeastern Brazil. Why did you choose that area?
Nicolelis: Yeah! That is a very important part of the whole project is the idea that by taking science or a big scientific project to a part of Brazil that was underdeveloped and that had not received the attention that other regions of the country had in terms of scientific investments, we could actually test this notion that science can be a driving agent of transformation; not only economic transformation like we know here in the United States and Europe, but also social transformation—that the values, the ethical, the ludic values that scientists apply to their daily life could actually be used to drive a whole educational program, a whole healthcare program and even a self-sustainable economic model that has science at the core. And the values of science [are in] the philosophy and that will, you know, help to find the scaffolding of a development plan. So it is in experiment, it is almost like a sociological experiment: How could science go there, get roots, and in working with the community, help develop series of fundamental issues like education and healthcare to a community that had not been in the radar screen, you know, of science in Brazil?
Soares: So, not just a self-contained institute where science is going on, but where it really actually affects the community and benefits the community.
Nicolelis: Yes, the idea is an open institute, an institute that has porous walls where the knowledge that is being
generating [generated] inside can easily migrate to the outside world and reach the public education system, the healthcare system and the economic development planning of the region; and transform the region in[to] a magnet for knowledge-based initiatives and that involves everything. You know, for instance we're going to have the first school in Brazil—and probably one of the first in the world—where kids go to school before they are born. Their mothers go to school, they return to school and when their kids are born, they enroll automatically in a fellowship that we're calling fellowship for life. They get a fellowship to be in that school from the early age all the way to the end of high school; we've a very empirical science-based curriculum guiding the whole education process of these children.
Soares: And the goal is not just to, maybe, train some future scientists, but also to have that way of thinking that basically benefit them, no matter what they do in life.
Nicolelis: Absolutely. The ambition is not to create a factory of scientists. Actually the goal is to allow these children to become critical thinkers and to develop their potential [in]
and whatever [their] potential is. In fact, we are creating a school where kids they have, you know, normal health and kids that are disabled in anyway study together, they are in the same classroom and they're basically developing at their own pace to reach whatever potential and dreams they have, whatever they are. So we just want to instill in them this idea that you can touch the impossible, you know, you can really pursue it, you may not even get there, but you're actually the pursuer of it; is [it's] a great adventure, whatever you want to do in life.
Soares: You're just giving them the opportunities.
Nicolelis: Yes, giving [them] the tools and the opportunity to flourish.
Soares: The next stage in the process: You want to see a biotechnology park built around this area.
Nicolelis: Yeah, in our attempts to develop a self-sustainable business plan for this project, we thought that the first major step after the construction of the institute and the establishment of social programs was the elaboration of a plan in which science, not only the science that is produced in the institute, but high[-end]
in science could be used as a way to generate wealth that could sustain all these activities, the basic research and the social programs. So, we are working in close collaboration with the Brazilian government in the state Rio Grande do Norte, where Natal is located, to elaborate a whole set of bills that will create a free enterprise only in that area, which is strategically located; you know, Natal is on northeast, close to the equator, at the tip of Brazil, [the] northeast tip of Brazil. So, it's the closest point to Europe, [the] United States and Africa, so [the] federal government of Brazil is investing, creating a huge cargo airport and a harbor there. So, what we want is to create really a serious science park, [a] free-enterprise zone, in which you can bring companies, both big and small, to produce scientific products for exploring. And surrounding this, you can create a service park that can allow a bidirectional interaction between the free enterprise zone and the City of the Brain of Natal in a way that you actually allow the community to benefit from the creation of such a mega-industrial structure. So, it's almost like a research triangle park here in North Carolina, but with a social mission, something that is rare to see in and around the world.
Soares: That is the difference. There are quite a few cities or, you know, communities that have been built by governments based on technology and science. But that is the difference in this one,
-- is the connection to the community, building a community at the same time.
Nicolelis: Yes. And basically allowing the community around, you know, Natal and the outskirts of the city to take advantage of that both economically and socially; you know, as I said, a bilateral interaction, where we can show to society that science has many more benefits than IP in creating a business. You know, you really can use science for many other things. One of our goals once Natal is built is to clone it, to clone
Soares: Now pretty soon, you're going to be addressing some very prominent audiences: the World Economic Forum in Davos in January.
Nicolelis: At Davos, I think I'm going to talk primarily about the model of doing science—how you can today really have a global lab and not only one lab, but how institutes in countries can belong to networks that can maximize tremendously the resources available to produce science; and basically connect the best minds in a field or in a discipline almost instantaneously to work together in projects that can really transform a society. And I think that's the future of science—
-- isa global science, is a science without frontiers. This is already part of the history of science, of course scientists perhaps would first want to propose the globalization as a good thing, but now we can see clearly that this may be the best business model to do science. So I want Davos to present the idea of the network of institutes in Brazil and how they could be translated into a network of institutes around the world, which we are already starting in neuroscience; but [it] could be much bigger than that.
Soares: Good luck with all of it.
Nicolelis: Thank you; thank you very much.
Steve: Christine Soares' article on Miguel Nicolelis and the efforts to create science cities is called "Building a Future on Science"; it's in the February issue of Scientific American and at our Web site. The Web site
in [and the] February issue also include[s] a commentary by the president of Brazil, Luiz Inácio Lula da Silva, co-authored by Nicolelis and Fernando Haddad about Brazil's plans to promote national development through education, especially science and technology education. That opinion piece is called "Brazil's Option for Science Education". And for more on the world Economic Forum, which starts next week in Davos, Switzerland, just go to www.weforum.org And check out Larry Greenemeier's January 15th news article on the monkey/robot research: It's called "Monkey Think, Robot Do" and it's at our Web site, SciAm.com
Now it is time to play TOTALL……. Y BOGUS. Here are four science stories; only three are true. See if you know which story is TOTALL……. Y BOGUS.
Story number 1: A new study shows that the lower your cholesterol levels, the easier it is for you to put on muscle mass while exercising.
Story number 2: An octopus in an English aquarium has a new favorite toy: Mr. Potato Head.
Story number 3: Also from England, Manchester is considering allowing the heat of bodies burning in crematoriums to be used to heat the rooms where families and friends of the dearly departed gather.
And story number 4: People who tasted the same wine in two different glasses thought one tasted better: the one they were told cost more.
Time is up.
Story number 4 is true. When people thought a wine was more expensive, they rated it as better tasting than the exact same wine that they thought didn't cost as much. Brain scans found that the taste center of the brain acted the same in both cases, but the pleasure center lit up more with [the] allegedly more expensive wine. For more, check out the January 15th episode of the daily SciAm podcast, 60-Second Science.
Story number 3 is true. They are considering using crematorium heat to also warm up the family members of the deceased. The hot gases that usually are released into the environment would be captured to heat up radiators and leave a slightly smaller carbon footprint.
Story number 2 is true. Louis, the giant Pacific octopus at Newquay's Blue Reef Aquarium in the U.K. loves to play with Mr. Potato Head. According to BBC News, the 18-month[-old] octopus apparently likes the colors, shapes, and movable parts of Mr. Potato Head; Louis may outgrow Mr. Potato Head though. The giant pacific octopus is the largest octopus species and can be more than 30 feet across.
All of which means that story number 1 about how low cholesterol levels make it easier to gain muscle mass is TOTALL……. Y BOGUS. Because a study from Texas A&M University found that lower cholesterol levels can actually reduce muscle gains from exercise. A study was done with men and women in their 60s. Those with higher cholesterol levels gained more muscle mass and strength. The researchers think that cholesterol's contribution to inflammation, which is bad around the heart, may be good for helping muscles recover from a workout.
Well that's it for this edition of the weekly SciAm podcast. You can write to us at podcast@SciAm.com and check out numerous features at the new SciAm.com Web site, including "Ask The Experts," blogs and opinions and slide shows and videos. For Science Talk, the weekly podcast of Scientific American, I'm Steve Mirsky. Thanks for clicking on us.