MIT researchers have found that a sugar-cube-sized piece of the brain helps prevent us from being lost in space.
In a study to be published in the April 9 issue of Nature, Associate Professor Nancy Kanwisher and postdoc-toral fellow Russell Epstein of the Department of Brain and Cognitive Sciences found that a part of the brain called the parahippocampal cortex is involved in perceiving the local environment. The parahippocampal cortex gets very active when people look at photographs of indoor or outdoor scenes, but not when they look at photos of faces or objects.
This reaction, which is just as strong to a grassy field as to a furnished room, seems to be tied to our ability to perceive our surroundings and know where we are.
"When you walk into a room, you don't spend five minutes trying to get your bearings," said Dr. Epstein. "These processes happen very quickly and automatically, much like the response we see in the parahippocampal cortex when people view scenes."
This is not the first time a single part of the brain has been linked to a specific task. Professor Kanwisher, who has been using magnetic resonance imaging (MRI) for several years to study how we perceive faces, objects and scenes, says the results are surprising "in that few scientists would have predicted that this particular process--perceiving the layout of the local environment--would have its own special-purpose bit of brain dedicated to it."
Using MRI, researchers can precisely pinpoint brain activity in response to a stimulus. MRI produces pictures of "slices" of the brain with brightly colored areas indicating regions where neural activity is greater in one spot than in another.
The parahippocampal cortex feeds information to the hippocampus, which is in charge of establishing memories and is involved in encoding spatial memory in rats. "Our results could have consequences for how spatial memory is set up," said Dr. Epstein, who earned his PhD in computer vision, the study of how to create machines that "see."
The area of the brain that the researchers named the parahippocampal place area (PPA) reacts consistently to complex scenes such as open fields, city streets, store interiors or empty rooms. The only thing the images have in common is that they are all places in which people could imagine walking around.
When subjects were shown pictures of places such as a Somerville street lined with cars and triple-deckers, empty rooms and rooms filled with furniture, the PPA "lit up." The brain's reaction was just as strong to an empty room as a room full of furniture, but pictures of faces, furniture alone or distorted scenes that no longer represented a coherent space elicited a negligible or weak response.
The PPA response was strong and consistent in all subjects, even though the images they saw flashed before their eyes for a mere 300 milliseconds.
There was no difference in intensity of reaction when the subjects were shown familiar and unfamiliar scenes and places. The researchers wanted to make sure that they were not simply eliciting a memory response.
Dr. Epstein said he suspects the function of the PPA is to help us navigate through the world by providing a rough-and-ready representation of our immediate environment--for example, by figuring out where the walls of a room are.
It makes sense that humans have an innate ability to "know where we are in the world," he said. "Babies and rats use information aboutthe shape of their environment to orient themselves, and we think the PPA might be the part of the adult human brain that extracts this information.
"To me, this is a fascinating subject," said Dr. Epstein, who pointed out that it's extremely difficult to get a computer to "see" with any semblance of the ease and accuracy with which humans take in and process visual information. "It's really incredible that we have this ability to look at our surroundings and almost immediately know where we are."
This research is sponsored by the National Institutes of Mental Health and the Human Frontiers Foundation.
A version of this article appeared in MIT Tech Talk on April 15, 1998.
