Scientists Created a Video Game That Makes People Hear Sounds That Aren't Real

A new study induced hallucination-like perceptions in humans and mice to better conditions such as schizophrenia.
Image: Westend61 via Getty Images
Image: Westend61 via Getty Images
ABSTRACT breaks down mind-bending scientific research, future tech, new discoveries, and major breakthroughs.

Scientists induced people and mice to report hearing sounds that did not actually exist while playing a computer game, an effect called “hallucination-like perception,” according to a new study. 

The cross-species experiment sheds new light on the neural circuits that cause hallucinations, which are a common symptom in psychotic disorders such as schizophrenia.

A project led by Katharina Schmack, a research investigator at Cold Spring Harbor Laboratory on Long Island, NY, concluded that their results offer “a promising entry point to identify novel treatment targets” in the human dopamine system as well as “urgently needed mechanistic treatments for schizophrenia,” according to a study published on Thursday in Science. The team was led by Adam Kepecs, a professor of neuroscience and of psychiatry, and a BJC Investigator, at Washington University School of Medicine, where the experiment was conducted.


Schmack, Kepecs, and their colleagues opted to model hallucinations in humans and mice to combine the benefits of research with both species. Humans can speak about their subjective experiences, while mice have limited communication, which makes humans better subjects for psychiatric researchers. However, mice can undergo a wider array of neurobiological tests—in this study’s case, taking ketamine doses or dopamine boosts—that are logistically and ethically problematic to conduct in humans.

“The big challenge we are addressing here is really bringing this together by doing the same thing, both in mice and in humans, and using a model of hallucinations to bridge the subjective experience of hallucinations with the neurobiology, which we can study in mice,” Schmack said in a call.

To prepare for the experiment, Schmack and her colleagues trained dozens of mice to operate a computer game that played specific auditory cues, which varied in volume, against a background of messy noise. The mice learned to poke their noses into ports to report whether or not they’d heard the cue, and were given water rewards when they were correct.

The 220 people who participated in the study also played a version of the game in which they were asked to click their responses with a cursor, indicating if they thought they had heard a sound. When the mice or the humans reported hearing a cue that was not actually played, the event was noted as a hallucination-like percept (HALIP). 


Certain permutations of this basic experiment yielded interesting results: For instance, both mice and humans reported more HALIPs when the cues were played at a higher frequency, which the researchers think is due to a heightened expectation of hearing signals.

Interestingly, human participants who said that they experience spontaneous hallucinations in a questionnaire reported more HALIPs during the experiment, hinting at the existence of neural circuits that could make these experiences more likely in certain individuals. 

To assess similar circuits in mice, Schmack and her colleagues ran trials in which some of the rodent subjects took ketamine doses or dopamine boosts before they played the game. Both variations resulted in the mice reporting more HALIPs than in trials with no ketamine or extra dopamine influence. Given that the dosed mice showed no performance problems or other erratic behavior, the team concluded that they had isolated the role of ketamine and dopamine in increasing the hallucination-like experiences in mice.

The results support the well-corroborated association of hallucinations with dopamine, an important chemical messenger in the nervous system. It’s no surprise that excess dopamine is a key driver of these sensory experiences, as many treatments for hallucinations prioritize dopamine-blockers. But the cross-species nature of the study revealed new details about the specific processes that occur in the striatum, a part of the forebrain, that appeared to be related to the HALIPs.


“No one could really explain what is the process that mediates this,” Schmack said. “This is the new thing here that we can say: ‘Okay, it's really striatal dopamine, and we think it has to do with perception. It has to do with the way perception relies on expectations.’ That’s the big point here.”

In other words, dopamine in the striatum may moderate the balance between what we expect from our sensory system, and the real input we receive from the world. In the example of the computer game, the high frequency of auditory cues set up the expectation, in mice and humans, that the sounds should be anticipated, which caused the subjects to feel more confident that they had heard them, even when they hadn’t. When the nervous system contains too much dopamine, it shifts the balance toward the expectation of signals, upping the odds that an individual will experience them as hallucinations.

The team is optimistic that the study will help medical researchers devise more targeted treatments for people who suffer from conditions like schizophrenia, which often cause auditory hallucinations.

“We can go in all directions; we can look at all sorts of all regions in the brain and all different cell types, and we can really try to get better treatments,” Schmack concluded. “I think that's the excitement that we are feeling about this approach.”