In dark rooms, surrounded by illuminated screens, figures scuttle, duck, and navigate through twisting corridors and mazes. This isn’t the elaborate setup of a teen gaming lair; rather, it’s a high-end virtual reality system built for neuroscience’s favorite mammals: rodents.

Tour a neuroscience lab fifteen years ago and you might see rats or mice swimming in circular water tanks or exploring T-shaped physical mazes as part of research on memory, spatial navigation, and vision. By studying how rodents perform on these tasks, researchers try to understand similar brain functions in humans, which in turn could help find treatments for conditions like Alzheimer’s disease, ADHD, and autism. Today, though, a growing number of labs around the world are turning to virtual reality as a new experimental approach to answer the same questions about the brain.

Inside a Virtual Reality Lab

Have you ever been to a show at a planetarium, sitting inside a dome viewing images of the night sky? Aman Saleem, a Professor of Systems Neuroscience and Sir Henry Dale Fellow at University College London, says VR setups in neuroscience labs are “like planetariums for mice” – in which one mouse can move around and even control the projected images. By interacting with a simulated environment projected onto a physical dome, a rodent receives visual sensory inputs. These inputs activate sensory neural circuits in the brain, which in turn activates other circuits. Researchers then study the resulting brain activity through techniques like electrical recordings and fluorescence microscopy.

Depending on what brain regions are being studied, labs create different VR environments for their rodents. Saleem, who studies how vision is used to navigate space, places mice in planetarium domes on free-moving treadmills. These mice see down a long, black-and-white VR corridor and their movements control the scene. “If the mouse runs forward,” Saleem explains, “it goes forward in the corridor. If it walks backwards, it’s going to go backwards in the corridor – just like in the real world.” 

The mouse’s head is held fixed so that Saleem can view real-time electroencephalogram (EEG) recordings, which measure electrical activity in the brain. From electrodes implanted in its primary visual cortex and hippocampus, images are received from the eyes to create a “map” of the hallway for navigation and movement. Saleem hopes this will help him understand how vision and navigation are linked.

Mayank Mehta is the director of the Keck Center for Neurophysics and Professor of Physics and Astronomy, Neurology, and Electrical and Computer Engineering at the University of California, Los Angeles. The simulated environments found in his lab reflect his interests in memory and learning. Mehta uses VR to investigate questions such as “how does the healthy brain learn? How does the brain abstract out concepts and remember things? And what goes wrong in conditions like Alzheimer’s and ADHD?” 

To study this, Mehta invokes abstract idea creation in rats, through assigning them one of the most complex tasks they can do: “We ask the rat to solve a complicated maze in VR.” Rats are placed on a moving sphere inside a planetarium dome and, given a little incentive, they learn quickly to perform tasks like navigating to spotlights. “They have to figure out, what is the shortcut? Where is the reward? What is the bad place? What is the good place? When they do the right thing, they get a drop of sugar water, which they love,” Mehta says. “Surprisingly, neurons in hippocampus show new signals, never seen before in rodents, but very similar to those in the human brain. This means that rodents in VR illuminate human brain function or pathology. A great tool for pharmaceutical companies!”

It’s important to Mehta that his rats learn in a stress-free environment so that his findings reflect the activity of a normal, healthy brain. Instead of restraining his rats’ heads, he fits them with “tuxedo vests.” “They happily walk around all over the lab in their tuxedos,” Mehta says. “It mimics how their mothers pick them up from the back. When they don’t want it, they can still slip out of the vest. And they take naps in VR. That’s the gold standard, where they feel comfortable enough to fall asleep.” And his setup minimizes stress too. “When the rats look somewhere,” Mehta explains, “they can see their own paws, they can see their whiskers, they can see their shadows. There is no dizziness.”

Pros and Cons

Neuroscientists who use VR say the control it provides in experiments is invaluable. In traditional physical maze tasks, rodents move freely and respond to many sensory stimuli that are difficult to track. In VR though, “we can easily manipulate the visual environment and more easily understand how we go from vision to memory,” Saleem says. 

VR is also compatible with experimental techniques that can study the brain on a finer level. VR environments restrict the movement area of animals and sometimes also restrain their heads, allowing researchers to use instruments like intracellular electrodes one micron in diameter.

“We can look at the transformation across different brain areas and track how information is processed,” Saleem says. “So we can do research that is otherwise not possible.”

However, rodent behavior in VR may not track with behavior in real life. Like humans who know a racing game at the arcade isn’t real because it lacks the sensory data of real life, rats in VR mazes are aware that their experience doesn’t sync with smell, feel, and sound. Research from the Mehta lab has shown that the hippocampal neurons of rodents, which are involved in space perception, fire differently when the animals are in VR simulations as opposed to similar environments in the real world.

Whether the “virtual” aspect of virtual reality is a detriment depends on the research question being asked. For his research on learning and memory, Mehta believes that VR actually improves the predictive power of animal models.