It’s the dream of interrogators everywhere, from grade-school principals to CIA agents — being able to peer into a person’s head to see what he or she is thinking.
It sounds like the stuff of science fiction. But a team of researchers at Cornell University appeared to have glimpsed that dream when they confirmed what 19 young volunteers were thinking in a series of experiments.
Their findings, published last month in the journal Cerebral Cortex, relied on functional magnetic-resonance imaging, also known as fMRI, to delve into the thoughts of their research subjects.
The researchers eventually see the technology enabling health professionals to gain a better understanding of autism and advertising companies to see the private thoughts of consumers.
The process today requires hugely expensive hardware and insane amounts of data that takes researchers months to compile, he said.
“Computationally, this was a very big project,” Spreng said. “We were recording 200,000 data points every three seconds, and each volunteer was in the fMRI for 45 minutes.”
The experiment involved placing volunteers in an fMRI and asking them to picture four bus passengers, each with certain personality traits. Researchers then asked the volunteers to explain how the four passengers reacted when new people boarded the imaginary bus, which was full.
The results of the scans, displayed in colorful blobs, reflected how the brains of the research subjects responded when they thought of the unique personality traits of each of the four imaginary bus passengers.
FMRIs use large magnets to obtain detailed images of brain activity by measuring second-by-second changes in blood-oxygen levels.
“When we remember, we’re not just replaying a video in our heads,” said Spreng. “The brain reconnects disparate pieces of memories together. It’s a physical process.”
Imagining works almost the same way.
The massive volume of data gathered, coupled with the complex algorithms used to process it, led to patterns, such as: This blob means Subject A is picturing Bus Passenger 2.
Using the tools and techniques employed in the experiment for real-time applications, such as detective work, seems many years away, however.
“I don’t see this being used for interrogation anytime soon,” said Spreng. “Our experiment was 100% [controlled] and even then there was so much noise in the data. Even moving your head causes noise.”
Spreng also acknowledged a potential for marketers.
If Facebook has taught us anything, it’s that every scintilla of thought is worth something to someone. Not that we need Cornell researchers to tell us what our “friends” are thinking.
“Does anyone hold back any thoughts in social networks?” Spreng joked. “Who needs an MRI?”
Jim Nash is an award-winning business, tech and science journalist whose work has appeared in The New York Times, The Economist Group and Scientific American.Tags: Business,Government,Healthcare,Technology