It’s often called the mind’s eye.
“I can look at an object in the world around me, but I can also close my eyes and imagine the object,” says Varun Wadia, a brain scientist at Cedars‑Sinai Medical Center and Caltech. That ability to conjure faces, navigate with a mental map or picture familiar things is visual imagination.
Wadia and colleagues report in Science that imagined and perceived objects activate the same neurons and share the same neural code. “This has not been demonstrated before at the neural level,” says Kalanit Grill‑Spector, a psychology professor at Stanford who was not involved in the study. She says the finding could help build computer models of vision and disorders like macular degeneration, and guide development of sight-restoring prosthetics.
Thomas Naselaris, a neuroscientist at the University of Minnesota, notes the result helps explain how imagination supplements visual input. Objects present multiple views but we intuitively model unseen parts—imagining a car’s bumper or assembling familiar parts into novel hybrids like a unicorn.
The study examined 16 patients with epilepsy who already had electrodes implanted so doctors could localize seizure foci. That setup let researchers record activity from more than 700 individual neurons while participants viewed images on a screen. The images spanned categories—faces, animals, plants, words—and included small objects such as sunglasses and water bottles.
The team focused on neurons in the ventral temporal cortex, a region involved in object recognition. They recorded which neurons fired to each image and how often, letting them decipher the code neurons use to represent visual information.
In a second phase, participants closed their eyes and imagined objects they’d previously seen while the same neurons were monitored. “About 40% of those neurons reactivated when you were imagining the object,” Wadia says, “and they reactivated with roughly equal strength.” The overlap was so consistent that researchers could tell which specific object a subject was imagining—an airplane, for example—and infer details such as its size, angle and whether it was inside or outside.
This single‑neuron evidence supports earlier brain‑imaging findings that seeing and imagining recruit overlapping circuits, but imaging methods like fMRI cannot resolve individual neurons. The work also builds on prior studies, including those by Doris Tsao, showing how visual systems recognize faces and objects.
The study does not explain conditions such as aphantasia, where people cannot voluntarily form mental images. Rutishauser, who led much of the experimental work, recounts a scientist telling him, “I don’t see anything when I close my eyes.” People with aphantasia may rely on words or concepts rather than imagery; understanding their neural strategies will require direct neuronal studies.