Neural subpopulations in marmoset area MT/MTC detect and discount saccade-related retinal motion
This study reveals that in marmoset area MT/MTC, distinct neural subpopulations integrate both corollary discharge and visual motion cues to either suppress or discount saccade-induced retinal motion, thereby supporting the perceptual stability of the visual world during eye movements.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
The world does not move when we look at it, yet our eyes are constantly jumping from one point of interest to another. These rapid jumps, called saccades, happen several times every second, sweeping our vision across the landscape at speeds that can reach 700 degrees per second. If the brain simply recorded every shift of the image on the retina, the world would appear as a chaotic blur of motion. Instead, we perceive a stable, continuous scene. To achieve this, the visual system must actively filter out the motion caused by our own eyes. This filtering process, known as saccadic suppression, is a fundamental trick of perception that allows us to navigate without being overwhelmed by the blur of our own movement.
Scientists have long debated how the brain achieves this stability. One theory suggests the brain uses a "copy" of the eye movement command, a signal sent from the motor planning centers to the visual areas to tell them to ignore the incoming blur. Another theory proposes that the visual system simply reacts to the massive, wide-field motion that occurs during a jump, using that visual chaos itself to trigger a suppression mechanism. A new study on marmoset monkeys has now looked inside the brain to see how these two possibilities play out in the areas responsible for processing motion. By recording the activity of hundreds of individual nerve cells while the animals looked at natural scenes, blank screens, and even in total darkness, researchers found that the brain uses a combination of both strategies, but with a surprising division of labor among different types of neurons.
The researchers focused on two specific regions in the marmoset brain, areas known for their expertise in detecting motion. They implanted tiny recording devices to listen to the electrical signals of individual neurons while the animals performed a simple task: looking freely at images on a screen. To understand what was driving the neurons, the team varied the conditions. Sometimes the animals looked at rich, natural photographs. Other times, they looked at a blank gray screen where no visual motion could occur during an eye jump. In a third condition, the lights were turned off completely, and the animals made eye movements in total darkness. Finally, the researchers simulated the visual effect of a saccade by moving the image on the screen while the animal's eyes remained perfectly still, allowing them to separate the visual motion from the actual eye movement.
The recordings revealed a diverse population of neurons, each reacting to eye jumps in a different way. About 39 percent of the cells fired a burst of activity almost immediately after an eye movement began. These "early response" neurons were particularly interesting because they seemed to know exactly which direction the eye had moved. This directional tuning held true whether the animal was looking at a complex image, a blank screen, or even in the dark. This suggests that these cells receive a direct signal from the brain's motor command centers, a kind of internal memo that says, "We are moving the eyes to the right," allowing the cell to anticipate the motion before the visual blur even arrives.
However, the story is not just about receiving a memo. These same early neurons also fired vigorously when the image on the screen was moved to simulate a saccade, even when the animal's eyes were still. This indicates that they are also highly sensitive to the visual motion itself. The researchers found that these cells likely integrate both the internal motor signal and the external visual motion to detect that a saccade is happening. They appear to be a fast-acting detection system, using the internal command to get a head start and the visual motion to confirm the event.
In contrast to these early detectors, another group of neurons behaved differently. These cells did not fire when the eye moved; instead, they went quiet. This suppression was most pronounced in neurons that were tuned to see fine details and high-contrast patterns. When the eye jumped, these cells shut down, effectively turning off the visual channel to prevent the brain from processing the blur. This group of cells was biased toward having broader electrical waveforms, a signature often associated with the brain's primary excitatory neurons, the ones that send information to other parts of the brain.
Perhaps the most striking discovery was a small subset of neurons that acted as a filter for reality. These cells responded strongly when the image on the screen was moved to simulate a saccade, but they remained completely silent when the animal actually made a saccade. They could tell the difference between motion caused by the world moving and motion caused by the eye moving. The researchers called these "saccadic omission" units. Because they are the ones that relay information to the rest of the brain, their ability to ignore self-induced motion is likely the key to why we see a stable world. If these cells reacted to the blur of an eye jump, our perception would be disrupted every time we looked around.
The study also mapped these different cell types to specific layers within the brain tissue. The fast-acting, early response neurons were found mostly in the deeper layers and the input layers, where they appeared to be inhibitory cells that could quickly dampen activity. The cells that ignored the motion of real saccades were found mostly in the superficial layers, which are the output layers that send signals to other brain areas. This arrangement suggests a specific circuit: the fast, inhibitory cells detect the eye movement and trigger a suppression signal, while the output cells in the upper layers use that information to filter out the noise, ensuring that the final image sent to our conscious perception remains steady.
By combining recordings from natural viewing, blank screens, and darkness, the researchers were able to disentangle the complex mix of signals that create our stable visual experience. They found that the brain does not rely on a single mechanism to cancel out the blur of eye movements. Instead, it employs a coordinated team of neurons. Some act as rapid detectors that combine internal commands with visual cues, while others act as filters that specifically ignore motion caused by our own actions. This division of labor allows the brain to maintain a clear, stable view of the world, even as our eyes dart around at incredible speeds.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.