A new study finds mental imagery can reactivate many of the same neurons used for visual perception, reshaping how scientists explain imagination.
The way we picture the world in our mind often feels effortless, like flipping through a private gallery. But inside the brain, researchers are now uncovering the circuitry that makes that mental trick possible.
In a study published April 9 in Science. Misryoum reports that imagining an object can reactivate at least part of the neural activity normally involved in actually seeing it.. The work addresses a long-standing question in cognitive neuroscience: do perception and imagination share the same building blocks at the level of individual neurons—or do they merely recruit the same brain regions?
Mental images mirror perception at the neuron level
To test that idea. Rutishauser and colleagues studied 16 adults with epilepsy who already had electrodes temporarily implanted in their brains for seizure localization.. While researchers recorded activity from more than 700 neurons in the ventral temporal cortex—a key area for representing visual objects—participants viewed hundreds of images spanning five categories: faces. text. plants. animals. and everyday objects.
The recordings revealed that a large fraction of neurons were “category-responsive.” Machine learning analysis then refined the picture further. showing that most of these neurons weren’t just responding to broad categories; they were selective to particular visual features within the images.. In other words, the neural code in this region carried specific information about what people were seeing.
When participants imagine, some of the same code lights up
Next came the crucial step: can mental imagery reuse that code?. The researchers focused on a subset of participants and asked them to conjure images of objects they had previously viewed.. When participants performed that mental task. about 40% of the neurons that were active during perception showed similar activation patterns during imagery.
Misryoum notes that the team didn’t stop at “activation similarities.” Using the neural data. they reconstructed the pictures participants were asked to recall. offering a functional readout of what the brain’s activity pattern contained.. The overlap between perception and imagery was therefore not just anatomical or regional—it appeared to reflect meaningful representational content.
A “generative model” for imagination
Taken together. the findings support the idea that the brain uses a generative model: a system that can rebuild internal predictions from stored neural patterns.. Rather than imagining being a purely separate stream of activity. it may work by reactivating elements of the perception-driven neural code.
That matters because many earlier theories assumed overlap between seeing and imagining. and functional imaging studies often suggested shared brain regions.. But overlap at the region level doesn’t guarantee overlap at the neuron level.. This study moves the argument forward by providing evidence that individual neurons involved in perception can. in substantial numbers. re-engage during imagination.
A researcher not involved in the work described the results as something the field had been waiting for—highlighting how difficult it is to answer this kind of question with earlier methods.. With direct recordings from hundreds of neurons. Misryoum frames the study as a step toward closing the gap between “similar brain areas” and “shared neural constituents.”
Why this could matter beyond neuroscience curiosity
Mental imagery is more than a parlor trick.. People navigate spaces, plan actions, read and understand stories, and create art by manipulating internal representations.. If the same neural machinery partly underwrites both perception and imagination. then disruptions to that machinery could plausibly affect more than just “visual thinking.”
Misryoum also points to a clinical angle raised by the authors: disturbances in mental imagery may contribute to conditions such as schizophrenia and PTSD.. Those disorders involve symptoms that can include altered perception and internal experience.. While the study doesn’t diagnose anything. it provides a mechanistic foundation that could help future research identify where and how imagery processes break down.
What’s next: beyond simple objects and toward complex imagery
The investigators studied relatively controlled, object-based imagery—faces, text, animals, plants, and everyday items.. That leaves an open question: do similar mechanisms scale up to more complex forms of imagery. like imagining a new artwork that hasn’t been seen before. or dreaming up a scene with multiple interacting elements?
Rutishauser suggests a testable hypothesis, but Misryoum reads the current work as a starting point rather than a final answer. Complex imagination likely depends on broader networks, including memory, attention, and executive control, not just ventral temporal representations.
Still. the study offers a clear direction: treat imagination less as a misty mental process and more as a computation that can. at least in part. reuse perception-linked neural representations.. If that framing holds. it could reshape how researchers design therapies—especially those aimed at improving or re-tuning the brain’s internal models.
For now, the brain’s ability to “see” without eyes looks less like magic and more like biology—an internal replay system built from the same cellular patterns that handle the real world.
Venus vs Jupiter: The Solar System’s Weirdest Planet
Artemis II reentry: NASA crew says they’ve done their homework
Space Tracking Follows Amputee Sea Turtle on Return to Ocean