Individual neurons differentiate between internal and external motion.
Neuroscientists have identified how the brain differentiates between visual motion in the external environment and motion caused by an observer’s own movement. This challenge, known as the “motion-source separation problem,” has long puzzled researchers. Now, for the first time, scientists have uncovered the precise mechanisms behind this distinction.
A study published in Cell details how researchers at the Sainsbury Wellcome Centre (SWC) at UCL developed an innovative experimental approach to isolate key components of locomotion. Their findings reveal that individual cells in the primary visual cortex of mice integrate motor and vestibular signals to determine whether retinal visual flow results from external motion or the animal’s own movement.
“Every day we take for granted that we know whether we are moving or something is moving around us. But no one knows how the brain does this. We wanted to design an experiment that would allow us to solve this motion separation problem,” said Professor Troy Margrie, Associate Director at SWC and lead author of the study.
The Translocator: A Novel Experimental Setup
Together with engineers in the FabLab at SWC, the team developed a unique new system called the Translocator. This experimental setup consists of a passive treadmill that mice can choose to run on, while watching screens displaying a virtual moving corridor. The entire treadmill apparatus is also physically moved forward along a rail, synchronized with the speed at which the mouse chooses to run.
“We built on the principles of virtual reality setups, where an animal runs on a treadmill while being shown visual flow that is coupled to its movement. But in addition, we added translation in the forward direction, so that animals could actually experience locomotion (i.e. moving from A to B) according to their own running speed. This is why we called it the Translocator,” explained Dr Mateo Velez-Fort, Senior Research Fellow in the Margrie Lab at SWC, and first author on the paper.
The Translocator consists of a passive treadmill that mice can choose to run on, while watching screens displaying a virtual moving corridor. The entire treadmill apparatus is also physically moved forward along a rail, synchronized with the speed at which the mouse chooses to run. Credit: Sainsbury Wellcome Centre
This experimental setup allowed the team to isolate the fundamental elements of locomotion. For example, the researchers recorded the speed profile of a mouse actively running over 1.2 meters. They then placed the animal back at the start and replayed the same speed while blocking the treadmill, so the mouse was being passively moved rather than actively moving. This allowed the team to obtain a pure vestibular signal that was identical to the combined running and vestibular signal.
The scientists also obtained a pure motor signal by letting the mouse run on the treadmill while keeping the overall apparatus stationary, so that the mouse wasn’t translated.
“The Translocator setup allowed us to get a pure motor signal, a pure vestibular signal, and combined motor and vestibular signals. This meant that for the first time, we were able to pull these things apart,” explained Professor Margrie.
Neural Responses in the Visual Cortex
Using Neuropixels probes, state-of-the-art electrodes for simultaneous neural recording, the researchers recorded from the primary visual cortex and observed that approximately 50% of cells and particularly those in deep layers 5/6 responded to visual flow, running and translation.
“We wanted to know if this convergence of inputs was a general rule in the cortex, and so we also recorded from other areas, including the somatosensory cortex and the retrosplenial cortex, in darkness. We found that the motor and vestibular signals converge in many places in the brain, so this seems to be a fundamental property of the organization of many cortical areas,” explained Dr Velez-Fort.
It was previously thought that sensory representations had to be sent to other parts of the brain to be integrated with internal cues used for navigation. In contrast, the researchers at SWC found that primary sensory areas in the cortex have immediate access to the internal motion status of the animal.
Surprisingly, the team also found that the activity recorded from neurons in the primary visual cortex was very similar for both a natural and unnatural scenario. The same amount of neural activity was observed when animals were running and being translocated, as when mice were running but not being translated forward. This led the researchers to propose that running must suppress translation input. They tested this theory using a mathematical model developed in collaboration with Professor Claudia Clopath, which they found to support this phenomenon. The model also predicted that if the running speed was not coherent with the actual speed of the head, then an error would be signaled by the vestibular pathway. This prediction was then verified by additional experiments.
This work shows that many cortical areas including primary sensory areas are constantly being updated and receiving feedback from other modalities. In the case of the vestibular system, it is used to generate an online internal reference frame to provide context regarding the motion status of the observer.
Reference: “Motor and vestibular signals in the visual cortex permit the separation of self versus externally generated visual motion” by Mateo Vélez-Fort, Lee Cossell, Laura Porta, Claudia Clopath and Troy W. Margrie, 19 February 2025, Cell.
DOI: 10.1016/j.cell.2025.01.032
This research was funded by the Sainsbury Wellcome Centre core grant from the Gatsby Charity Foundation (GAT3361) and Wellcome (219627/Z/19/Z) and a Wellcome Trust Discovery grant (214333/Z/18/Z).
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.