Scientists just discovered long-sought-after 'grandmother neurons'

What happens in your brain when you recognize your grandmother? In the 1960s, some neuroscientists thought a single brain cell called the "grandmother neuron" would light up only at the sight of your grandmother's face. Almost immediately, neuroscientists began to dismiss the theory — a single neuron could not correspond to one idea or person, they argued. 

More than 50 years later, new research in monkeys shows that "grandmother neurons" may exist after all. In a study published on July 1 in the journal Science, researchers found a small area of the monkey brain that responds only to familiar faces. Up to three times as many brain cells in this area responded to familiar faces than to unfamiliar ones. The study follows research showing that certain parts of the human brain respond to specific categories, including one region primarily dedicated to faces. One study even found that individual neurons in different parts of the brain responded only to specific celebrities and landmarks. But few studies had found any part of the brain that reacts specifically to personally familiar faces. 

Though the new research did not identify individual cells devoted to a single person, the brain cells the researchers found share some crucial qualities with the theorized "grandmother neuron."

"In some sense, you can say they are grandmother neurons," said Winrich Freiwald, a professor of neuroscience and behavior at The Rockefeller University in New York City, who led the new research. "They have this unique combination of vision and memory."

The researchers examined the temporal pole, a poorly understood area near the bottom of the brain that Freiwald and lead study author Sofía Landi, a postdoctoral fellow at the University of Washington in Seattle, had identified as one of two areas that might be involved in familiar face recognition in a study published in 2017 in the journal Science. (The previous research was completed while Landi was a doctoral student in Freiwald's lab.) 

For the new study, the researchers used functional magnetic resonance imaging (fMRI) to scan the brains of two rhesus monkeys while they looked at images of monkey and human faces mixed with some other images. The brain scans served as a guide so the researchers could place electrodes in two areas of the monkey's brains — one in the temporal pole and one in another area of the brain that responds to faces in general but that past research suggested would not necessarily distinguish between familiar and unfamiliar ones. 

These electrodes enabled the researchers to monitor the activity of individual brain cells in the two areas. Brain cells in both areas lit up when the monkeys were shown images of monkey and human faces. But only the temporal pole cells distinguished between personally familiar and unfamiliar ones: When the monkeys saw photos of their pals and relatives, those temporal pole cells lit up three times more than when unfamiliar monkey photos were shown. These neurons barely responded to other faces, including both familiar and unfamiliar human faces as well as unfamiliar monkey faces. 

The discovery goes against prevailing wisdom in neuroscience. Generally, scientists believe that diverse areas of the brain must communicate with each other to process information. But this research indicates that "it's one area, this region, and it's there for this one purpose — recognizing people we know," Freiwald said. "That's amazing." 

The researchers also obscured the images of faces to varying degrees to see how the brain responses might differ. In the generic face-processing area, more cells gradually responded to the images as it became clearer and clearer that they were faces.But the response from temporal pole cells was different. They responded very little to highly blurred images, but as soon as the clarity reached a certain threshold, many of the neurons responded all at once to familiar faces. The researchers think this effect corresponds to the "a-ha" moment of recognition of a familiar face, say of your grandmother. 

When the researchers measured how fast the cells responded, they were surprised to find that there wasn't much difference between the two areas. The general face-processing area, which seems to engage only if an image is a face, responded to faces in about the same time as cells in the temporal pole area responded to only familiar faces. That is "very, very surprising," said Freiwald, because the assumption was that the grandmother neurons would take longer, since the person would first have to identify the image as a face, then associate it with a long-term memory of a specific person. 

The new research, though in many ways groundbreaking, comes with limitations. It was done on monkeys, not humans, and only on two individuals. However, Freiwald notes that rhesus monkeys, as highly social primates, are the best animal models to use for a study like this and are thought to have very similar face-recognition processing to humans. 

The researchers also don't know how exactly face information is being sent to this temporal pole region. The temporal pole doesn't directly process vision or store long-term memory, and because there aren't known pathways between the temporal pole and these other parts of the brain, the route that information might take to get there is still unknown. 

The insight could eventually help people who can't recognize others, said Freiwald. For instance, people with dementia and those born with prosopagnosia, or "face blindness," sometimes can't recognize close friends or even family members, something Freiwald imagines is "crushing."

Freiwald also noted that recognizing a person isn't a purely visual, or even sensory, experience. "There's also almost an emotional quality to it, like 'Huh, I know this person,'" he said. "That is ignited, we think, by this area, but there must be so much more involved in it."

Originally published on Live Science.