Scientists Hack Brain Perception With Lasers, Creating Optical Illusions

Lab Mouse Study Shows Brains Can Be Tricked Into ‘Seeing’ Shapes That Don’t Exist

BERKELEY, Calif. — Your brain does more than record what your eyes see. It fills in the gaps, using past experiences to make sense of incomplete information. A new study by scientists at the University of California, Berkeley and the Allen Institute in Seattle demonstrates this process: scientists recreated optical illusion-like activity in mouse brains using laser light and without showing the animals any actual pictures.

The findings show that some brain cells in the visual cortex do not just copy what enters the eyes. Instead, they predict missing details, creating signals that represent shapes not physically present. This supports the idea of predictive coding: the brain constantly guesses what is likely to be there and updates those guesses when new information arrives.

How Optical Illusions Expose the Brain’s Hidden Predictions

Optical illusions are a simple way to test this. Take the Kanizsa triangle: three black circles with wedges cut out can make you “see” a bright triangle, even though none is drawn. In this study, researchers used that illusion with mice and found special brain cells that fired for the fake triangle but not for the separate black circles.

When scientists activated these cells with lasers, the mice’s visual cortex produced the same patterns it would have if the animals were viewing the illusion. The brain built the representation of a triangle from inside itself, without input from the eyes.

This shows that illusions are not just tricks. They reveal something deep about how perception works: our brains do not wait for perfect information but jump ahead to make the best guess.

Two Kinds of Neurons Work Together to Build Vision

The study, published in Nature Neuroscience, identified two groups of brain cells:

• Segment responders: cells that honestly report the parts of the image, like the edges of the black circles.
• IC-encoders: cells that fire for the illusory triangle, not the pieces. They receive more signals from higher brain areas and appear to be more attuned to the broader picture.

Essentially, segment responders are like eyewitnesses, describing exactly what they saw. IC-encoders are like detectives, piecing together the missing parts of the story.

When scientists switched on IC-encoders with light, the activity spread and made the brain act as if a triangle was there. Turning on the segment responders did not have the same effect.

This finding illustrates how different roles within the same brain area interact: some neurons focus on the evidence, while others attempt to make sense of the entire scene.

Auto-Complete for Sight

Think of how your phone suggests words while you type. The brain does something similar with vision. When an image is incomplete, whether it’s because of shadows, objects blocking each other, or quick glances, the brain fills in what it thinks is missing.

The team showed that this happens in different layers of the visual cortex. One layer (layer 4) mostly reports raw eye input. Another (layer 2/3) shows the illusion effect, blending input with prediction. This division explains how the brain balances real data with smart guesses.

That separation may explain why illusions feel so convincing. Part of the brain insists on the raw evidence, while another part builds a prediction that can override what is missing.

Why Optical Illusions Trick Both Humans and Animals

Everyday vision is often incomplete. Objects hide, light changes, and motion is fast. Predictive brain cells allow us to make sense of partial information, like recognizing a friend’s face through a fence or spotting a shape in the dark.

The Kanizsa illusion tricks not only humans but also monkeys, mice, fish, and even insects. This suggests that filling in gaps is a shared feature across many animals. That makes sense; evolution favors quick guesses when survival depends on spotting prey or predators in an instant.

The study also provides clear evidence for predictive coding, showing that specific neurons can be switched on or off to trigger illusion-like activity. This makes prediction a built-in feature of how the brain sees, not just a theory.

Understanding these circuits could have broader implications. It may one day help explain disorders where perception goes wrong, such as hallucinations in mental illness, or guide new technologies that mimic how the brain interprets complex, messy information.

The Broader Picture

The discovery also highlights the significant advancements in neuroscience tools. Not long ago, scientists could only measure small groups of neurons at a time. Now, with advanced probes, imaging, and light-based stimulation, they can both watch and control thousands of neurons while an animal is awake and processing the world.

This ability to both record and manipulate brain activity opens up new possibilities. Researchers can now ask not just which neurons respond to illusions but whether turning them on can create the illusion-like state in the brain. In this case, the answer was yes — and that is a major step forward.

It also suggests that other senses may work in similar ways. Just as vision fills in missing shapes, hearing may fill in missing sounds, and touch may fill in missing textures. Predictive coding may be a general rule of how brains build reality from incomplete inputs.

While powerful, this research was done in mice. We do not yet know how directly it applies to human vision. The experiments focused on one type of illusion and on primary visual cortex, leaving open whether similar prediction mechanisms exist for other illusions or in other brain regions.

Another unknown is development: do IC-encoders emerge through experience, or are they hardwired? And do different individuals rely on predictive coding to the same degree? These questions remain for future work.

What’s clear is that perception is not a passive recording of the world. It is an active construction. The brain fills in missing pieces, sometimes accurately, sometimes not. By pinpointing the neurons that carry predictive signals, this study shows how brains turn fragments into the seamless pictures we experience every day.

Disclaimer: This article is for general informational purposes and should not be taken as medical or professional advice.