University researchers in Pittsburgh pursue breakthrough in understanding mind
You understand this sentence, but no one's really sure how.
Though the brain allows people to perceive and understand the world, it remains one of the least understood things in it. Scientists know the mechanics — how neurons fire and which chemicals flow. And they understand what emerges — the blend of behaviors, quirks and traits that make a personality.
The mystery is what happens in between.
“It's like this cavern that exists. There's just this entire black box where magic appears to happen, and we don't understand how it works,” said Marlene Behrmann, psychology professor at Carnegie Mellon University.
Unlocking the secrets of how we think could revolutionize treatments and teaching methods for stroke victims, children with reading disabilities and people with a host of neurological disorders. The concentrated mix of computer scientists, psychologists and medical researchers inhabiting the laboratories at Carnegie Mellon and the University of Pittsburgh could make Oakland the site of the next breakthrough.
“I think we're heading for the perfect storm, and when that storm takes place, we'll be in the optimal position,” Behrmann said. “There's so much going on here simultaneously. It feels like we're in the Wild West.”
Carnegie Mellon's incoming president, Subra Suresh, emphasized the crossover between technology and human behavior during a recent interview with Tribune-Review reporters and editors. Advances in weather prediction, for example, do no good if people don't react to the warnings, he said.
“It's not only how tech influences daily life, but how human behavior in conjunction with technology influences daily life,” Suresh said.
Understanding how the brain works can help businesses improve.
“It can be as crass as better marketing. It can be as meaningful as creating better cars so we don't kill 50,000 people every year,” said Michael Tarr, a Carnegie Mellon psychology professor who will become department chairman on Wednesday. “It is so fundamental to so much of what we do as a society.”
That's why the university's psychology department includes computer scientists such as professor David Plaut. He uses computers to simulate processes of the human brain. His computer models learn to read in a simulation that isn't nearly as complex as a human brain but is complete enough to show what happens when one part of the network breaks down.
“We can have a stroke in our (computer) model,” Plaut said.
Strokes, caused by a lack of blood flow, can kill sections of a brain. They affect nearly 800,000 people a year, according to the Centers for Disease Control and Prevention. About 1.7 million others endure traumatic brain injuries a year, according to the CDC. The injuries emerged as a leading medical problem among troops wounded in Iraq and Afghanistan.
Plaut's brain-like computer network re-creates these traumas without damaging a person's brain. “If you have a good theory of how a system operates and breaks down, you can be much better at fixing it,” he said.
Eventually, Plaut hopes that what researchers learn about brain function will help children with developmental disabilities. If they know how a child without the disability learns to read, and they understand how a certain disability affects the brain, “you could tailor (education) to individual kids' strengths and weaknesses.”
Without such understanding, doctors stumble around the dark cavern that Behrmann described.
“We can't treat diseases of perception, or brain injury or dyslexia or autism, unless we have a good understanding of the processes that are being injured,” Tarr said. “If you've ever known anyone who had a concussion, the treatment is basically a lot of little physical activities. They have no idea why they work.”
Behrmann, Plaut and Tarr focus much of their research on the mysteries of vision, including trying to get a computer to understand what humans seem to know naturally.
The human eye “is measuring a million points of light. Each (point) has some level of brightness and wavelength. From that, you need to distill surfaces, materials, what surfaces go with which objects — and you need to do that under varying conditions,” Tarr said.
Imagine standing in a living room lit by bright lamps on opposite sides of the room. If one lamp goes out, the light changes dramatically. Shadows form. Colors shift. The overall light gets dimmer. “You've changed every point of light on your retina,” Tarr said.
Yet a person who just turned off the light, even a child, can recognize the room.
“I was always amazed at the things that people can do, particularly the things people can do quickly and seemingly effortlessly,” Plaut said. When programming a computer, the “things people can do easily — opening their eyes and looking around and understanding everything they see — turned out to be the hardest problems.”
Trying to bridge psychology and computer science could lead to devices that hold up parts of the brain that don't work on their own, Behrmann said.
For example, even though scientists don't know how the brain recognizes faces, they can program computers to do it.
Devices such as Google Glass — a computer with a camera, small screen and earpiece — could someday handle facial recognition for people with prosopagnosia, a psychological disorder that renders people unable to recognize faces, Behrmann said.
It wouldn't be the first time people used technology as a shortcut, she said.
“Airplanes fly,” she said. “They don't fly like birds, but they fly.”
Mike Wereschagin is staff writer for Trib Total Media. He can be reached at 412-320-7900 or firstname.lastname@example.org.