How Your Brain Creates Depth Perception

Jenn Hoskins
8th August, 2025

How Your Brain Creates Depth Perception

Miniscope calcium imaging of the primary visual cortex in freely moving mice (a–g) revealed that neuronal firing rates and activity levels were comparable between passive visual cliff and active depth-discrimination tasks (h–j), providing a stable baseline for investigating context-dependent depth processing.

Image adapted from: Dong et al. / CC BY (Source)

Key Findings

  • Researchers at Southern Medical University found that the brain's primary visual cortex (V1) has distinct groups of neurons for passively observing depth and actively making decisions based on it
  • These V1 neurons encode objective positions in the environment, not just distance from the animal, and even anticipate future spatial relationships for planning movements
Our ability to navigate and interact with the world relies heavily on perceiving its three-dimensional (3D) structure. This complex process, known as depth perception, allows us to judge distances, recognize objects, and plan our movements. While it is known that the brain's primary visual cortex (V1) plays a role in processing visual information, its specific contributions to different aspects of depth perception, particularly in active, naturalistic settings, have been less clear. Understanding how V1 processes depth is crucial for unraveling the neural architecture behind our spatial awareness and decision-making. Recent research from Southern Medical University[1] has shed new light on how the primary visual cortex (V1) contributes to depth perception. This study investigated V1's role in both passive and active depth-related tasks, aiming to determine if distinct groups of neurons within V1 are responsible for different types of depth processing. The findings indicate that specific populations of V1 neurons are indeed selectively active depending on whether an animal is passively observing depth or actively engaging with it. Furthermore, the study revealed that V1 neurons tend to encode objective positions in space rather than just the animal's egocentric distance (distance from oneself), and that V1's processing of egocentric distance appears to be anticipatory or "prospective." To explore these questions, the researchers used freely moving mice and employed in vivo calcium imaging, a technique that allows scientists to observe the activity of individual neurons in the brain of a living animal. Mice were engaged in two main types of tasks: a "visual cliff" task, which is a passive scenario where the animal perceives a drop-off without needing to act on it, and an "active depth discrimination" task, where mice had to make decisions based on perceived depth, such as judging a gap. This approach allowed the researchers to observe how V1 neurons responded to different demands on depth processing. The study's findings build upon and expand several earlier lines of research. Previous work has shown that our perception of 3D shape involves the combination of various depth cues, like binocular disparity (the slight difference in images seen by each eye) and perspective. Studies using fMRI (functional magnetic resonance imaging) in humans, for instance, found that higher visual areas, such as hMT+/V5 and the lateral occipital complex, are more involved in reflecting perceived 3D shape from these combined cues, suggesting that these areas hold 'combined-cue' representations, rather than early visual areas like V1[2]. The new study, however, highlights V1's fundamental role in processing different types of depth information, potentially providing the initial signals that higher areas then integrate. The use of freely moving mice in the Southern Medical University study is particularly significant, as it reflects a more naturalistic context for studying brain function. This aligns with earlier research demonstrating that in natural settings, sensory processing and motor output are closely linked. For example, a study on freely moving mice showed they could accurately jump across gaps using vision, even relying on monocular cues (depth cues available to one eye, like motion parallax or position parallax) when binocular vision was unavailable[3]. This earlier work also found that the primary visual cortex was crucial for this distance judgment under both binocular and monocular conditions, directly supporting the current study's focus on V1 in active, depth-related tasks. The observation that mice in the earlier study altered head movements under monocular conditions to gather more depth information further underscores the importance of active engagement with the environment for spatial perception. Moreover, the new findings that V1 neurons prefer encoding "objective positions" and are involved in "spatial navigation and decision-making" resonate with prior research on V1's role in integrating various signals. It has been shown that the primary visual cortex integrates visual stimuli and motor feedback, with neuronal activity in V1 being shaped by experience. In tasks where mice learned to locate a reward in a virtual environment, V1 neurons became responsive to the expected reward location. Crucially, when visual cues were absent, both behavioral and neuronal responses relied on self-motion-derived estimations, but when visual cues were available, they became the primary driver[4]. This suggests that V1 is not just a passive receiver of visual input but actively encodes behaviorally relevant spatial locations, adapting its processing based on available cues and self-motion feedback. The "prospective" nature of V1's egocentric distance discrimination, as found in the Southern Medical University study, also connects with the concept of internal models in brain function. Generative models propose that the brain uses internal representations to predict sensory input. Previous research in mice has shown that with experience in a virtual environment, V1 neurons become increasingly informative of spatial location, even exhibiting responses that predict upcoming visual stimuli[5]. This predictive activity, along with strong responses to omitted expected stimuli, suggests that V1 forms an internal representation of the visual scene based on spatial location and compares this with incoming visual information. The current study's finding that V1's egocentric distance processing is prospective further supports this idea, implying that V1 is not merely reacting to current depth information but is actively anticipating future spatial relationships, which is vital for planning movements and navigating complex environments. In summary, the research from Southern Medical University provides valuable insights into the versatility of the primary visual cortex. By demonstrating that V1 contains functionally segregated neuronal populations for passive and active depth tasks, encodes objective positions, and processes egocentric distance prospectively, this study significantly advances our understanding of V1's crucial role in spatial navigation and decision-making. It ties together previous findings on V1's involvement in naturalistic behaviors, its integration of self-motion and visual cues, and its capacity for predictive coding, painting a more complete picture of how this fundamental brain region contributes to our perception of the 3D world.

HealthAnimal Science

References

Main Study

1) Characterization of depth perception information inferred from neuronal activity in primary visual cortex

Published 7th August, 2025

https://doi.org/10.1371/journal.pone.0329788

Related Studies

2) 3D shape perception from combined depth cues in human visual cortex.

Journal: Nature neuroscience, Issue: Vol 8, Issue 6, Jun 2005

3) Distance estimation from monocular cues in an ethological visuomotor task.

https://doi.org/10.7554/eLife.74708

4) The Impact of Visual Cues, Reward, and Motor Feedback on the Representation of Behaviorally Relevant Spatial Locations in Primary Visual Cortex.

https://doi.org/10.1016/j.celrep.2018.08.010

5) Experience-dependent spatial expectations in mouse visual cortex.

https://doi.org/10.1038/nn.4385


Related Articles

An unhandled error has occurred. Reload 🗙