The perception of time is a peculiar sensory experience: it is constructed without dedicated sensory receptors or specialized neural pathways. Nevertheless, time is intrinsic to all sensory perceptions. Any stimulus that reaches our senses - for example, the notes of a melody - unfold over time and is accompanied by a subjective sense of temporal flow. Yet, despite its fundamental role, time has rarely been studied as a sensory feature.
Through a series of psychophysical and 7T fMRI experiments, we investigate the link between sensory coding (e.g., the coding of visual space or speed) and the neural representation of time (see, for example, Centanino, Fortunato, Bueti 2024).
Like other sensory modalities, time perception is not a faithful representation of the physical world. Instead, it reflects an interaction between incoming sensory information and prior expectations about the structure of the environment, shaped by past experience and statistical regularities.
Using a combination of psychophysics, high-resolution fMRI, and pupillometry, we investigate how environmental statistics and participants’ expectations influence both the subjective perception and the neural coding of time.
Stressful events are widely reported to distort the perception of time. Empirical research shows that threatening or arousing stimuli often lead to duration overestimation, suggesting that time is experienced as passing more slowly under stress (e.g., Droit-Volet & Meck, 2007; Lake et al., 2016). However, the neurobiological mechanisms underlying this effect remain poorly understood.
This project investigates whether cortisol - a key hormone in the human stress response - contributes to stress-related distortions of time perception. We employed a mixed, double-blind, placebo-controlled pharmacological design, administering hydrocortisone or placebo to healthy male participants. Stress-related physiology was monitored using salivary cortisol measurements and pupillometry.
Participants completed two psychophysical tasks assessing whether cortisol influences: (i) duration perception, measured with a duration reproduction task; and (ii) low-level visual processing, measured with an orientation discrimination task.
This study was conducted in collaboration with the Emotion and Memory Lab directed by Prof. Ulrike Rimmele (CISA, University of Geneva) and was supported by the European Union’s Horizon Europe Program (Marie Skłodowska-Curie Actions), awarded to Dr. Monika Riegel (Project ID 101064781: Neuromodulation of episodic memory – how stress influences time in memory and perception).
We investigate the neural correlates of time perception using behavioral experiments combined with electroencephalographic (EEG) recordings. To address the high dimensionality of EEG data, we employ multiple methodological approaches that enable the extraction of meaningful spatiotemporal features.
Our work includes the application of classical techniques - such as principal component analysis (PCA), hidden Markov models (HMM), and independent component analysis (ICA) - as well as the development of innovative methods to characterize neural dynamics associated with temporal estimates.
In collaboration with Alessandro Laio’s Data Science Group (SISSA), we apply causality-detection algorithms based on the Information Imbalance of distance ranks to test: (i) whether changes in neural activity causally relate to perceived stimulus duration across participants, and (ii) how information flows between EEG channels. This framework allows for robust inference of causality in human EEG data using a purely variational approach.
In humans, the processing of visual duration engages a distributed network of brain regions, including primary visual cortex (V1) and supplementary motor area (SMA). However, the functional roles of these regions in temporal computations remain unclear.
A simple hypothesis proposes that V1 encodes stimulus duration at the sensory level, while SMA - positioned higher in the processing hierarchy - decodes this information for task-related decisions. We test this hypothesis in two transcranial magnetic stimulation (TMS) studies, one of which combines twin-coil TMS with simultaneous EEG recording. These studies aim to characterize the directionality and temporal dynamics of communication between V1 and SMA during time perception.
Many real-world sensory signals evolve in a preferred temporal direction: smoke disperses, waves break, and people walk forward rather than backward. These processes reflect the arrow of time, and their degree of temporal irreversibility can vary.
We investigate whether human perceptual decisions are sensitive to such graded cues of temporal irreversibility. Using psychophysics and computational modeling, participants are tested with visual motion sequences generated by physically grounded non-equilibrium systems - specifically, drifted Brownian trajectories with controlled levels of temporal asymmetry.
By manipulating the strength of the arrow of time, we aim to determine whether decision formation relies solely on average motion statistics or whether observers also exploit time-asymmetric fluctuations characteristic of non-equilibrium dynamics.
This research is conducted in collaboration with Mathew E. Diamond (SENSEx Lab, SISSA) and Edgar Roldan (Stochastic Thermodynamics and Biophysics Group, ICTP), integrating perceptual decision-making, computational neuroscience, and stochastic thermodynamics.