The human brain can play tricks on us, even with our own memories, generating remembrances that we believe to be valid, even though they’re not.

But why do human have false memories, and where do they come from?

Recently a Penn research team led by Professor of Psychology Mike Kahana, Associate Professor of Neurology and Bioengineering Brian Litt (pictured) and lead author Per Sederberg, a former Ph.D. student at Penn now working at Princeton, sought to answer these and similar questions. The result of their collaboration is a new study that, for the first time, has identified the specific brain waves responsible for differentiating between real and false memories. The study is a breakthrough in the understanding of human memory.

“We were in a position to answers questions that could not be answered before,” says Kahana, senior author of the paper and director of Penn’s Computational Memory Lab. “Our data was unprecedented both in kind and in scope.”

The findings, based on the observed brain activity of candidates for neurosurgery, helps scientists understand how we successfully store and retrieve our memories. It may also lead to improved strategies for mapping brain functions, information that could advance treatments of several neurological conditions, including Parkinson’s disease, depression, obesity and obsessive/compulsive disorder.

The study could be used to help doctors reduce cognitive complications in patients undergoing surgery for epilepsy.

“This really opens the door to all sorts of treatment for disease,” says Litt.

To test whether certain brain waves could be used to distinguish true from false memories, 52 epilepsy patients were asked to study lists of common nouns which they were later asked to recall. Occasionally they made mistakes, recalling words that were not listed.

The participants had all been outfitted with intracranial electrodes to record the origin of their seizures, giving scientists the opportunity to monitor electrical activity in their brains during the recall game. What the researchers found was that a certain brain wave, known as the gamma rhythm, increased when participants studied a word that they would later attempt to recall. When the same fast gamma waves spiked in the half-second prior to participants correctly recalling an item, scientists concluded that gamma wave activity is a reliable predictor of whether a memory had and had not actually happened.

“We are able to measure neural activity in deep brain structures as people attempted to recall items in a memory test,” says Kahana. “That allowed us to determine what happens in the brain in the moments just prior to recall, when the mental process of remembering is taking place.”

In addition to helping scientists better understand the process of memory recall, the study may also lead to better surgical outcomes for patients with epilepsy.

In some cases, surgery to remove the malfunctioning area of the brain is the best option for stopping seizure activity.

But pinpointing the exact origin of the seizures can be problematic. When one area of the brain is damaged, functions will gravitate to surrounding, healthy areas.

“The brain is constantly building new bridges,” Litt explains. “In neurosurgery, one issue we face is finding a way to take out what needs to come out without interfering with important functions.”

But now, by understanding and mapping functional networks in the brain, researchers are developing better ways to help epilepsy patients retain cognitive ability after brain resection surgery. Litt’s team of bioengineers and neuroscientists is designing surgical implants that automatically map brain functions “to help us understand how brain circuits are organized and function.”

“The whole area of neurodevices is exploding at Penn and this is just the tip of the iceberg here,” says Litt.

Adds Kahana: “There is no doubt that we are at a very early stage in our understanding of the neural basis of human memory. Unlike sensation and action, which largely reflect the input and output of the nervous system, memory processes depend on understanding internal brain states whose character is still largely unknown.”