Credit: sommart sombutwanitkul on Shutterstock
In A Nutshell
- Scientists used a high-powered MRI scanner to show the brain may process brief durations through three distinct stages rather than relying on a single internal clock.
- Different brain regions handle different parts of the job: some encode how long something lasts, others organize that information, and forward regions categorize it as “short” or “long.”
- Activity in those forward regions correlated with each participant’s individual internal sense of where “short” ends and “long” begins, a boundary that varied from person to person.
- The study involved 13 adults and focused solely on visual timing, so whether the same system applies to sound or other senses remains an open question.
Catching a baseball, tapping along to music, or hitting the brakes at a yellow light: every one of these split-second judgments depends on a brain that can measure tiny slices of time with remarkable precision.
Scientists have long known that dozens of brain regions light up during timing tasks, but a harder question has lingered. What exactly is each region doing? A new study using powerful brain imaging technology has found a surprisingly organized answer. Rather than a single timing center running the show, the brain handles brief durations through a three-stage pipeline, with each stage carried out by a distinct set of regions.
Published in PLOS Biology, the research reveals that brief visual durations are processed through at least three distinct stages spread across the brain’s outer cortical surface. Some regions appear to encode how long something lasted. Others sort durations into organized categories. A third group does something more personal: it houses each individual’s own internal boundary between “short” and “long,” a boundary that varies meaningfully from person to person. That last finding is especially notable because it offers one of the clearest links yet between brain activity and how individuals judge time.
How Scientists Mapped the Brain’s Time Perception System
A research team from the International School for Advanced Studies in Trieste, Italy, recruited 13 healthy adult participants, six of them women, with an average age of about 30, for a brain imaging experiment. Each person lay inside a 7-Tesla MRI scanner, among the most powerful available and far stronger than machines used in routine medical care, while watching a small colored patch flash on a screen for varying lengths of time ranging from two-tenths of a second to eight-tenths of a second. Before scanning began, participants memorized a reference duration of half a second. Their job during each trial was simple: decide whether each flash was longer or shorter than that memorized reference.
Behind that simple task, the researchers were doing something technically demanding with the brain data. Using a modeling approach borrowed from research on how the brain maps visual space, they identified which duration each tiny patch of brain tissue responded to most strongly. Each cluster of brain cells has something like a preferred note on a piano, and the modeling approach identified that preference across cells spanning the brain’s entire outer surface. In total, the team identified 93 brain regions across both hemispheres that showed significant activation during the timing task.

Three Stages of Time Perception in the Brain
What emerged from the data was a clear division of labor. In visual areas at the back of the brain, neurons responded more strongly to longer flashes, suggesting a basic encoding of duration rather than a judgment about time. It functions as a raw tally of how much happened.
Moving into regions near the top of the head, roughly in the middle of the brain’s processing chain, the picture changed. Neurons there showed preferences spread evenly across the full range of tested durations, from shortest to longest. Cells tuned to similar durations also clustered together on the brain’s surface, forming organized maps not unlike the well-known maps for visual space or sound frequency found elsewhere in the brain. Researchers interpret this as a readout stage, where the brain has moved past simple encoding and is now representing each duration as its own distinct entity.
Forward-positioned regions, including areas tied to motor planning, speech processing, and internal body awareness, showed something different again. Rather than spanning the full range, neuron preferences in these areas clustered tightly around the middle of the tested range, close to the half-second reference duration participants had memorized going into the experiment.
Where Subjective Time Perception Lives in the Brain
That clustering alone was notable, but the most revealing finding came from what those forward-region preferences correlated with. Across several of these areas, the preferred duration of local brain cells tracked with each participant’s own internal cutoff between “short” and “long.” That cutoff varied from person to person. Participants whose internal boundary fell slightly longer than average also had neurons in these regions that preferred slightly longer durations, and vice versa. Critically, this correlation grew stronger moving from the back of the brain toward the front, suggesting that brain activity more closely tracks each person’s subjective sense of time at progressively higher levels of processing.
By looking at how preferences related across all 93 regions, the researchers also identified three main functional clusters. One included visual areas at the back, handling raw encoding. A second grouped middle and forward regions supporting duration readout and subjective categorization. A third brought together motor-related areas likely involved in the action side of timing decisions, specifically the physical response of pressing a button.
Motor and sensory areas near the groove dividing the front and back halves of the brain showed a preference for the shortest durations, but the researchers do not believe this reflects genuine timing. Activity in these regions before each button press was tied to how quickly participants responded, pointing to motor preparation rather than time measurement.
Altogether, rather than a redundant chorus of regions all doing the same thing, a clear processing chain emerged: encode, read out, categorize. At the end of that chain, in the folds of the brain’s forward regions, sits something that may reflect how each individual experiences the passage of time.
Paper Notes
Limitations
Only 13 participants took part in the study, a relatively small sample that may limit how broadly the findings can be applied across the general population. Brain imaging data were drawn from a previously collected dataset within the same research group, meaning the original experimental design was not built specifically around every question this study addressed. Because the task involved only visual timing, it remains unclear whether the same three-stage organization would appear with other senses, such as hearing. As with all correlational analyses, the observed links between brain cell preferences and individual perception cannot establish cause and effect on their own.
Funding and Disclosures
Funding came from the European Research Council under the European Union’s Horizon 2020 program (grant agreement no. 682117 BIT-ERC-2015-CoG) and from the Italian Ministry of University and Research under the PRIN22 call (project ID: 2022CCPJ3J), both awarded to researcher D. Bueti. Funders had no role in study design, data collection, analysis, or the decision to publish. No competing interests were declared.
Publication Details
Authors: Valeria Centanino, Gianfranco Fortunato, Domenica Bueti (International School for Advanced Studies, Trieste, Italy) | Title: Neuronal populations across the cortex underlie discrete, categorical, and subjective representations of visual durations | Journal: PLOS Biology, Volume 24, Issue 3 | Published: March 26, 2026 | DOI: 10.1371/journal.pbio.3003704 | Academic Editor: Christopher Pack, McGill University, Canada | Data Availability: Supporting data are available at https://doi.org/10.17605/OSF.IO/2TEQU. MRI data are provided in preprocessed format to comply with privacy regulations. Published under the Creative Commons Attribution License.







