A new study from NYU shows the organizational techniques our brains use to sort experiences based on their similarities.
The paper titled "Consolidation Promotes the Emergence of Representational Overlap in the Hippocampus and Medial Prefrontal Cortex" has been published in the journal Neuron and was authored by Lila Davachi, an associate professor in NYU’s Department of Psychology and Alexa Tompary, a recent NYU doctoral recipient.
The primary focus of the study was to explore how memories transform over time.
“It is as if to make sense of the world, the brain re-organizes distinct individual experiences into information clusters — perhaps signaling the emergence of conceptual knowledge,” observes Davachi in the study.
To achieve this, the neuroscientists at New York University conducted a series of experiments to determine how we process experience. Subjects were asked to learn a series of object-scene associations, they viewed several individual objects, such as a tennis racquet, on a computer screen.
Each object was paired with pictures of four repeating scenes, such as a beach scene. Using fMRI imaging, the scientists then tested the volunteer’s ability to match the tennis racquet with the scenes they saw at two time periods; both right after the experience and one week later.
It was during these recall periods that the researchers looked at the subject’s neural patterns of activation associated with memories.
The results found that soon after learning, there was no overlap in the pattern of activation associated with memories for the objects paired with the same scene picture.
Though, after one week, the patterns were more overlapping in the brain’s hippocampus (our brain’s memory storage system) and its medial prefrontal cortex (mPFC), meaning that the brain had filed the information according to their similarities.
"This aspect of the research points to the tension between 'good memory' and learning -- if we remember each individual experience as it was encountered, are we able to effectively learn about the underlying regularities across experiences?" asks Tompary in the study. "We see evidence for this competition in our neural analysis of memory structures in the brain."
The Power of Memory
Regarding the formation of these memories, a few years ago scientists at the University of Leicester were able to glimpse new memories being formed through fMRI imaging.
They noticed that single neurons, in the hippocampus region form new associations very quickly. Changes occurred at the exact moment of learning, even after a single presentation of an image.
"Think of the [cortex] as a huge library and the hippocampus as its librarian," wrote the prominent Hungarian neuroscientist Gyorgy Buzsaki in his 2006 book Rhythms of the Brain, as reported by the BBC.
Memory storage is something researchers aim to achieve with artificial neural networks, such as Google's DeepMind, which suffered from a phenomenon called “catastrophic forgetting” in which it would master a single task then call it quits--automatically wiping previous memories.
To tackle this, researchers at DeepMind and Imperial College London modeled an algorithm after the hippocampus that allows a program to learn on task after another, using knowledge gained along the way.
Cumulatively, this neuro research into the power of our ability to remember shows that the human brain is a hard act to follow.
Via: Science Daily