At this exact moment, your brain is busily performing an inconceivable number of tasks.
As you read this, it deftly turns symbols into words and words into thoughts. It’s regulating your breath, keeping track of your limbs and is orchestrating where they will move next. It feels pain or comfort, anxiety or pleasure. It recognizes that the room you’re sitting in now is too warm or too cold, and is cognitively situating you in the present, rather than the past or future.
And while grey matter may look unassuming (an over-steamed haggis comes to mind) it’s actually more complex than any supercomputer on the planet. Its roughly 86 billion neurons perform thousands of intricate connections, “speaking” to each other in a way no scientist fully understands.
Not that the complexity is stopping them. In the past year alone, massive brain mapping projects have been announced around the world that would give researchers new roads of insight into the human mind. In Europe, there’s the Human Brain Project, spearheaded by École Polytechnique Fédérale de Lausanne together with collaborators from 86 other European institutions. The project, which garnered a stunning 500 million euros from the European Commission, will use a giant supercomputer to simulate how the brain works by creating a neuron database. Then, in April 2013, U.S. President Barack Obama officially announced a brain-mapping initiative that would receive $100 million in federal funding in the first year alone.
Meanwhile, Canadian and German neuroscientists have released the first complete 3-D digital map of a post-mortem human brain. Consisting of a staggering terabyte of data, it goes beyond the macroscopic level and involved slicing up the brain of a 65-year-old once-healthy woman into over 7,000 segments, thinner than a strand of human hair.
This sudden excitement and interest in brain mapping is understandable. For the first time, our computers are hefty enough to handle the digital load. Imaging technology has also come a long way this past decade. The two combined conditions create a perfect storm for innovation. By mapping the brain, it’s expected that the data could lead to new inroads into understanding diseases such as dementia, depression or Alzheimer’s.
The Waterloo brain
Knowing the enormity of the task, one might think that simulating the workings of a human brain would be an impossible venture — pure science fiction. In fact, the word “impossible” is one that Waterloo professor Chris Eliasmith often heard when he first set out to write his book, How To Build A Brain (Oxford University Press, 2013), more than three years ago. He’s the director of Waterloo’s Centre for Theoretical Neuroscience and holds a Canada Research Chair in Theoretical Neuroscience, while being jointly appointed to the Philosophy and Systems Design Engineering departments and cross-appointed to Computer Science.
In short, if someone is going to ask the big questions about the mind, it’s Eliasmith.
He decided to write the book because he felt that the work coming out of his lab was at a point where he could give readers a high-level overview of what it was accomplishing. Some researchers were getting results around decision-making, others around memory, spatial navigation, emotional and social reasoning, and motor control. The overarching theme of the Centre is theoretical, or mathematical modelling, and researchers are using many similar methods and software to tackle their own problems and projects.
Could all of this consistent work be combined to create a massive, simulated, functioning model of the brain? He thought so, but the trick was to prove it.
“About eight months into writing the book, I thought, ‘No one is going to believe me unless we can actually show that we can take all of these parts, put them together and have a model that performs action, perception and cognition,’ ” he says.
Spaun is spun
Enter “Spaun,” or Semantic Pointer Architecture Unified Network, currently the world’s largest functional brain simulation, which actually emulates behaviours while also modelling the physiology that underlies them. Although it lives in a simulated environment in the Centre’s lab, it can count, remember lists of numbers, solve cognitive tasks and even find number patterns — no software programmer’s direction required.
Eliasmith explains that the word “functioning” is what sets his team’s work apart from the other large projects. Not only does it offer insight into how specific aspects of the brain work, but it can be used for practical purposes, too. For instance, there’s a hypothesis that as we age, people suffer cognitive decline as brain neurons die. It’s a pretty straightforward idea, but difficult to test on human test subjects as it’s unethical to randomly kill a healthy person’s brain cells just to see if the person then performs poorly on a test.
But with models like Spaun, his team can show that as neurons die, there’s a progressive decline in the ability to perform the tests no matter how “smart” the virtual person started out.
“With a model, they’re virtual neurons. It’s OK to kill them,” he jokes.
Emotion, meet brain
Paul Thagard, a professor of Philosophy and director of the Cognitive Science Program at Waterloo, agrees that to make any real headway into brain network research, turning to the computer is critical. Computers help to collect a vast amount of data, and provide a tool for developing models that can simulate how the brain works.
“If you want to understand the really interesting things about what the mind does, ranging from perception to language to consciousness, you have to build models,” he says.
Thagard was awarded the Killam Prize this year for his work as a leading thinking-theory researcher, and for his pioneering work in the philosophical use of computer models. He has spent the past 20 years looking at brain and emotion.
He says that the brain is not nearly as modular as once thought. In other words, there’s no such thing as a happy part of the brain, or a sad part. Instead, numerous regions of the brain communicate to create emotions. He’s interested in how people switch among them.
“When my kids were little, it was really striking with a three-year-old how fast they could go from being perfectly happy to miserable and back again,” he said. “So what’s going on there? What kind of neural processes can produce a shift in emotion?”
There are numerous practical reasons to use map data to study brains and feelings. Images created by functional magnetic resonance imaging, or fMRI, can go a long way. Comparing brain images of healthy people to those of depressed, anxious or schizophrenic patients could unlock secrets about how those mental illnesses manifest.
Yet, Thagard says, he’s just as interested in feeding his curiosity about the world.
“My motivation is theoretical,” he says. “The brain is just really fascinating. People have emotions and can solve complex problems. They have conscious experiences. I want to know why.”
Glass half gone
Here’s something else that’s fascinating: the effects of right parietal brain lesions on behaviour. In other words, when some people experience a stroke, these brain lesions form and can result in a condition called neglect.
The behaviour that goes along with neglect is both poignant and bizarre. Patients behave as though the left side of the world no longer exists. You place a plate of food in front of an afflicted person and he or she will eat only the food on the right side. (Turn the plate around and she believes you’ve given her more.) Patients dress only the right side of their bodies and men shave only the right side of the face.
“It’s not a laziness thing. They fully believe they’ve finished these activities and can’t represent that they haven’t,” says James Danckert, a professor in Psychology and Chair of Cognitive Neuroscience and Canada Research Chair (Tier II) in Cognitive Neuroscience.
Danckert says that “represent” is the operative word when it comes to his work with the human mind. He uses fMRI to create images to map people’s brains so he can try to find out what they mentally represent while interacting with the world. In other words, what they actually see in their mind’s eye so they make sense of their surroundings. It’s important research, considering that it’s estimated that 40 per cent of all right-hemisphere stroke patients will show neglect at some stage.
To figure out what is happening in the brain after a stroke, he first wanted to learn how a healthy brain creates representations. He placed university students in an fMRI machine, fitted them with LCD goggles, gave them a response button box and had them play “rock, paper, scissors” against a computer. Eventually, he changed the game so that the computerized opponent chose the same option, say, rock, most of the time. The students quickly figure out the new bias. Meanwhile, he recorded their brain images.
“What that task shows is that you’re able to update your representation of how the world is functioning. Our stroke patients can’t really do this,” Danckert explains.
Testing healthy people’s brains could eventually lead to finding ways to help stroke patients learn how to create mental representations again.
“If I don’t know how a normal brain ought to function, then I can’t properly characterize deficits that are a consequence of brain injury,” he says. “If I can’t characterize those deficits, then I can’t rehabilitate them.”
Images and movement
Richard Staines, associate professor of Kinesiology and Canada Research Chair in Sensorimotor Control, also uses imaging techniques as well as brain stimulation techniques to help stroke patients and those with brain injuries rehabilitate. In Canada, there are more than 50,000 strokes each year. His work will hopefully lead to advancements in repairing functional impairments after a stroke occurs.
Still, Staines calls himself “more of a basic human physiologist” who studies processes of sensory information and how that gets represented in the brain. He wants to know how the brain is able to adapt in order to modify what it’s representing so it can change it — especially when the brain is overloaded with multiple sources of information.
In one study, he took healthy students and hooked them up with stimulators that gave them a gentle zap to the tip of the index finger. Usually the zaps were of medium strength, but a few were at quarter strength. The participant’s job was to indicate when they felt the lower amplitude one.
In another trial, the students were asked to shift focus to their little finger. Surface electrodes were placed on a cap to be worn on the head so they could measure the results. The students were able to adapt to the new situation.
Later, a similar trial was conducted with older people. They showed they could perform the task, but had difficulty with attention.
“Clearly, as all of us age, the number and connections of neurons globally in the brain will slowly diminish,” says Staines. “One of the possible side effects is that your ability to learn or adapt to new skills very quickly will diminish, too. Not absent, but slower in the process. ”
Into the future
Bryan Tripp, a systems design engineering assistant professor, is also a researcher in Eliasmith’s lab. He wants to figure out how our brains handle visual motor control. The area may seem a little less glamorous than, say cognition, or mapping the brain of a person in love, or figuring out why teens crave risk, but understanding the connection between vision and movement is crucial. That’s because the same cranial hardware does similar work.
“Once we get that part figured out and it becomes accessible through modelling, then I think the really cerebral, far-out stuff will fall into place pretty quickly,” he says. “It’s a path to understanding how the brain works.”
Eventually, Tripp says, he can see a day when mapping the brain, in all its incarnations, could result in creating artificial systems that perform tasks as easily as humans.
Would it be a massive technological advancement? Absolutely. But don’t expect to see lifelike androids walking around any time soon. Even Obama, when discussing the U.S. brain-mapping project, threw around the word “century,” rather than “decade.” After all, there are roughly 100,000,000,000,000 neural connections that must be modelled, a number that dwarfs the Human Genome Project.
Tripp admits that he may never see the full impact of his research and that of the hundreds of scientists intent on mapping the human brain. Still, that realization doesn’t curb his enthusiasm.
“The potential applications are huge,” says Tripp. “If we figure out how the brain works, then it will be revolutionary technology that can be used in all kinds of ways.”
