Revolutionizing neuroscience research through immersive virtual environments for cross-species studies
Imagine trying to understand the complex symphony of the brain by listening to only a single, isolated instrument. For decades, this has been the challenge facing neuroscientists. To maintain rigorous experimental control, research often relied on simplified tasksâa rat pressing a lever or a monkey tracking a dot on a screen. While these studies yielded invaluable insights, they fell short of capturing how the brain operates in the dynamic, rich environments of everyday life. As one research team put it, this traditional approach is "hardly reflective of the highly dynamic and varied environment that the brain encounters in everyday life" 1 .
The ultimate goal is to study the brain in naturalistic settingsâas an animal forages, explores, and makes decisionsâbut this comes with a trade-off: a loss of experimental precision.
How can scientists track the exact timing of neural activity when an animal is freely roaming? Virtual reality (VR) emerged as a promising solution, creating immersive worlds that feel real to the subject while allowing researchers to control every variable. However, building these VR environments has traditionally required advanced programming skills, creating a significant barrier for many research groups. The solution? A powerful, accessible, and versatile new toolbox called DomeVR, designed to unlock the brain's secrets across speciesâfrom humans to monkeys to mice 1 3 .
Study brain function in environments that mimic real-world complexity while maintaining experimental control.
Create immersive worlds that feel real to subjects while allowing precise control of all variables.
At its core, DomeVR is an innovative virtual reality environment built using Unreal Engine 4 (UE4), a state-of-the-art game engine known for its photo-realistic graphics. What sets DomeVR apart is its mission to make this powerful technology accessible to neuroscientists, regardless of their programming expertise. "We chose to create our VR toolbox using Unreal Engine 4 (UE4)," the developers explain, citing its "visual scripting language that makes coding for non-programmers more accessible" 1 3 .
This means that researchers can now design complex, naturalistic virtual worlds using simple drag-and-drop elements, much like building a level in a modern video game. They can populate an eight-kilometer-long grassy field with foliage, create intricate mazes, or place various stimuli wherever needed, all through an intuitive graphical interface 1 .
Immersive virtual environments created with DomeVR allow for naturalistic behavior studies while maintaining experimental control.
However, a game engine alone isn't sufficient for rigorous science. DomeVR bridges this gap by incorporating essential research features directly into its framework 1 3 :
This combination of user-friendly design and scientific rigor allows researchers to create truly immersive environments where subjects can behave in instinctual ways while generating lab-quality data.
To see DomeVR in action, consider a groundbreaking 2025 study that investigated a profound question: Are the internal cognitive states that guide behavior similar across different species? The experiment brought together macaque monkeys and mice within the same DomeVR environment to find out 4 .
The setup was elegant in its simplicity. Both species were placed inside a custom-made spherical dome where a lush, virtual meadow was projected around them. Their task was a natural one: forage for food. In this digital meadow, two distinct leaf shapes appearedâa target and a distractor. The monkeys, manipulating a trackball with their hands, and the mice, running on a spherical treadmill with their feet, would navigate through the VR environment to approach the target leaf to receive a reward 4 .
This wasn't just a test of perception; it was a test of decision-making in a context that felt instinctual to both species. The DomeVR system seamlessly translated the movements of both the monkeys' hands and the mice's feet into smooth navigation through the same virtual world, demonstrating its remarkable cross-species adaptability 4 .
Species | Number of Subjects/Sessions | Input Device | Task Goal | Measured Outcomes |
---|---|---|---|---|
Macaque | 2 monkeys / 18 sessions | Trackball | Approach target leaf | Hit, Wrong, Miss, Reaction Time |
Mouse | 7 mice / 29 sessions | Spherical treadmill | Approach target leaf | Hit, Wrong, Miss, Reaction Time |
The true innovation of this experiment lay in what the researchers measured. While the animals were engaged in the virtual foraging task, cameras recorded their faces. Using deep-learning algorithms, the team tracked subtle facial movementsâeyebrow shifts in monkeys, nose twitches in mice, and ear movements in both 4 .
The critical analysis focused on a 250-millisecond window before the visual stimuli even appeared. At this moment, with no external task cues to react to, any facial expressions had to be generated by internal statesâfluctuations in attention, anticipation, or alertness that originated from within the brain itself 4 .
Macaque monkeys using trackballs to navigate the virtual environment while facial expressions are recorded.
Mice on spherical treadmills navigating the same virtual environment as the monkeys.
The researchers then employed a sophisticated statistical model called Markov-Switching Linear Regression (MSLR). This model analyzed the patterns of facial features to identify distinct internal states and predict how these states would influence the animals' upcoming behavior 4 .
The results were striking. The model, trained purely on facial expressions before stimulus onset, could reliably predict two crucial aspects of behavior:
The animal would react (reaction time)
It would make (task outcome)
Even more remarkably, the relationship between these internally generated states and task performance was comparable between mice and monkeys. Furthermore, "each state corresponds to a characteristic pattern of facial features that partially overlaps between species" 4 . This suggests that despite millions of years of evolutionary divergence, the fundamental ways that internal cognitive states manifest and guide behavior may be shared across mammals.
Species | Number of Features Tracked | Examples of Tracked Features |
---|---|---|
Macaque | 18 features | Eyebrow movement, pupil size, nose position, ear orientation |
Mouse | 9 features | Whisker pad position, ear angle, nose movement |
Creating an experiment like the cross-species foraging study requires a sophisticated integration of hardware and software. DomeVR functions as a cohesive toolkit, bringing together several specialized components, each playing a critical role in the research process 1 3 6 .
Component Name | Type | Primary Function in Research |
---|---|---|
Unreal Engine 4 (UE4) | Game Engine Software | Creates photo-realistic, interactive virtual environments via drag-and-drop interface and visual scripting |
Dome Projection System | Display Hardware | A spherical dome and projector system that fully immerses the subject in the VR world, suitable for non-human animals |
Stimulus Classes (Image, Movie, Grating, Mesh) | Software Objects | Pre-built, customizable objects (images, videos, patterns, 3D shapes) that can be placed in the environment as task stimuli |
Tracking Systems | Data Acquisition Hardware | Records subject behavior; can include eye trackers, motion sensors, and deep-learning-based video analysis (e.g., DeepLabCut) |
Logging & Synchronization System | Data Management Software | Precisely timestamps and records all experimental events and subject actions, allowing alignment with neural data |
Experimenter GUI | Control Software | Allows researchers to monitor live performance, start/stop tasks, and adjust parameters in real-time without interrupting the experiment |
These tools collectively transform a game development engine into a precision scientific instrument. The logging system, for instance, tackles a problem that would be negligible in entertainment gaming but is crucial for neuroscience: the inherent timing uncertainties in game engines. By solving this, DomeVR ensures that a neuron's firing can be correlated with a mouse's decision to turn left within a time window narrow enough to be scientifically meaningful 1 3 .
The dome projection, eye-tracking, and logging systems are available as independent modules that can be integrated into other UE4 projects.
This flexibility, combined with the system's accessibility, effectively "democratizes" virtual reality research.
Allows more scientists to ask bold questions about the brain by customizing experiments to their specific research needs.
DomeVR represents more than just a technical achievement; it signifies a paradigm shift in how we study the brain. By providing a bridge between the controlled environment of the laboratory and the rich complexity of the natural world, it allows researchers to observe the brain in a context that closely mirrors its evolutionary purpose. The success of the cross-species facial expression experiment is a powerful testament to this, revealing that the hidden cognitive states guiding a monkey's or a mouse's choices can be decoded from their faces and are more similar than we might have imagined.
Studies in complex, realistic landscapes to understand how brains map environments.
Investigations during dynamic, multi-sensory experiences.
Explorations of how conditions alter perception and behavior in virtual worlds.
DomeVR and tools like it are opening a window into the brain, allowing us to observe its intricate workings not in isolated snippets, but in the full, flowing tapestry of natural behavior.
The journey to understand the mind's inner workings is now taking place in worlds of our own creation, built with the very brains we seek to understand.