Developing a State-Dependent brain Model of Time Consciousness

The NeuroSchool PhD Program of Aix-Marseille University (France) has launched its annual calls for PhD contracts for students with a master's degree in a non-French university and for  international co-supervised PhDs.

This project is one of the proposed projects. Not all proposed projects will be funded, check our website for details.

State of the Art

The subjective experience of time—how we perceive the sequencing, duration, and persistence of events—has traditionally been linked to the activity of specific neuronal units that fire in response to temporal features of external stimuli. However, modern neuroscience increasingly favors a dynamical systems perspective, emphasizing that cognition and perception arise from collective neuronal activity across large-scale networks rather than the activity of isolated neurons. Despite this shift, an overarching framework that explains time consciousness as an emergent property of whole-brain dynamics remains elusive.

Moreover, traditional time perception studies largely assume that the brain passively processes stimuli, focusing on external triggers while neglecting the role of ongoing spontaneous brain activity. Recent research suggests that neuronal fluctuations prior to stimulus onset significantly shape perception, influencing whether and how a stimulus is consciously experienced. The re-interpretation of previous experimental results through the lens of spontaneous brain dynamics requires a unifying theoretical and computational framework. Developing whole-brain models that incorporate these principles will provide a crucial step forward in understanding time consciousness.

Objectives

This project seeks to advance our understanding of time perception by:

• Developing a theoretical framework that links time consciousness to whole-brain dynamics, emphasizing the role of ongoing large-scale neuronal events—such as neuronal avalanches and cell assemblies—in shaping the experience of time.
• Constructing a novel computational model that integrates both stimulus properties and the brain’s prior state to predict perceptual awareness.
• Investigating the dynamics of perception using this model to explore how spontaneous brain activity influences event sequencing, duration perception, and perceptual integration.
   

Methods

The project will introduce a temporal response function (TRF) that models perception as a function of both external stimuli and the brain’s intrinsic state. The TRF will be developed and validated as follows:

Modeling stimulus-dependent brain response: At a given time t, spontaneous brain activity (e.g., extracted from MEG resting-state data) is combined with an externally applied auditory stimulus. The stimulus is modeled as a square pulse with amplitude A in [0,1] and duration τ in [5ms,200ms]. The TRF takes both the stimulus and the prior brain activity as inputs and generates an expected brain response. If this response exceeds a threshold in frontal regions (global ignition), the stimulus is classified as a "HIT" (perceived); otherwise, it is a "MISS" (not perceived).

Tuning TRF parameters to match known experimental data: The model must reproduce established experimental findings, particularly the sigmoidal dependence of perception on stimulus duration and amplitude. When the stimulus duration is too short (e.g., 5ms), no ignition should occur. Ignition refers to a sudden and widespread activation of neural activity across the brain, particularly in the global neuronal workspace. This rapid amplification of signals is thought to underlie conscious perception, distinguishing conscious from unconscious processing. Conversely, for longer stimuli (e.g., 200ms), ignition should consistently happen. Similarly, weaker stimuli should be more frequently missed, while stronger stimuli should be perceived with high reliability. Similarly, the model can be trained to ensure that no ignition occurs when the brain exhibits sleep-like activity (from sleep data).

Validating the model and expanding its applications: Once the TRF is optimized, it will be applied to new experimental paradigms to test its predictive power. For example:

• Perceptual Integration: When two stimuli are delivered in rapid succession, the model should predict whether they are perceived as a single event or as two distinct occurrences, based on the ignition response.
• Brain State Dependence: The model will simulate different network states prior to stimulus onset (e.g., activity restricted to the default mode network vs. activity concentrated in the visual cortex) to investigate how different functional states influence perceptual thresholds and dynamics.

Expected Results

This project is expected to:

• A perspective paper that establishes a new theoretical framework linking time perception to large-scale brain dynamics.
• Develop a TRF model capable of predicting whether a stimulus is perceived based on the brain’s prior state and stimulus characteristics.
• Provide insights into how spontaneous brain activity influences perception, offering a potential explanation for variations in time perception across different mental states (wakefulness, attention shifts, sleep, etc.).
• Validate the model against known experimental results and generate testable predictions for future studies in cognitive neuroscience and time consciousness.

Feasibility

This project is theoretically and methodologically feasible within a doctoral research timeframe. It builds upon well-established neuroscience paradigms, such as predictive processing, dynamical systems theory, and whole-brain modeling. The computational modeling approach will leverage existing neuroimaging data (MEG, EEG) to train and validate the TRF, ensuring that model predictions align with empirical findings. The project’s interdisciplinary approach—combining neuroscience, cognitive science, and computational modeling—ensures that it aligns with ongoing advances in the field and can be completed using accessible experimental and computational resources.

Within Aix Marseille Université, NeuroMarseille brings together 8 research laboratories and NeuroSchool, a graduate school in neuroscience, to increase the attractiveness of the university, international collaborations, interdisciplinarity, links with the clinical and industrial worlds and the integration of students into professional life. 

Launched in July 2018, NeuroSchool unifies and harmonizes the training of the third year of the Bachelor of Life Sciences (Neuroscience track), the Master's and the PhD in Neuroscience. 

• Master's degree from a non-French university in neuroscience or related field
• Fluent in English

The ideal candidate should have:

• A background in computational neuroscience, cognitive neuroscience, or a related formal science (e.g., physics, engineering, applied mathematics).
• A strong conceptual interest in the neuroscientific and phenomenological aspects of time perception.
• Experience or strong interest in computational brain modeling and predictive processing.
• Proficiency in programming and data analysis (Python, MATLAB, or similar).
• Familiarity with neuroimaging techniques (MEG, EEG) and the analysis of spontaneous brain activity.