Conference Agenda

Sense of time is one of the most fundamental aspects of human consciousness. We present two studies that, for the first time, reveal the neural mechanisms underlying the influence of the heart on duration perception using electroencephalography (EEG), testing if and how the subjective experience of time arises from an intricate brain-heart interplay.

The EEG studies utilised our behavioural findings showing that perceived stimulus duration was distorted by autonomic signals arising from the heart, and that this temporal distortion was modulated by experienced arousal (Arslanova et al., 2023 Current biology). The first EEG study examined the neural correlates of temporal distortions when cardiac signals impacted emotionally neutral stimuli (i.e., participants judged the duration of visual Gabor patches), whereas the second EEG study focused on cardiac effects on duration perception under different levels of experienced arousal (i.e., participants judge the duration of faces showing neutral of fearful expression).

The first EEG study (N = 41) showed that cardiac signalling suppressed later stages of visual processing, which was correlated with contraction of perceived durations. The second EEG study (N = 48) revealed distinct mechanisms by which arousal and cardiac signals shape subjective duration perception – an early modulation by arousal, followed by a later modulation by cardiac signal.

Our results reveal how cardiac signals shape conscious experience by exerting top-down attenuation of sensory processing, how time may be intrinsic to sensory response, and how affective context drives the effect of the heart on our sense of time.

Over recent years, the critical role of body-brain interactions in the emergence of consciousness has gained increasing recognition. In this context, brain-heart interactions have been proposed as potential biomarkers of consciousness levels, with Heartbeat-Evoked Responses (HERs)—neural responses to heartbeat—being linked to perceptual consciousness, bodily self-awareness, and the presence of residual consciousness in brain-injured patients (1). HERs have been primarily studied using scalp EEG during wakefulness, where they have been associated with a widespread network of cortical regions (2). However, their dynamics during altered states of consciousness, such as sleep, remain largely unexplored. Additionally, the lack of standardized methodologies to distinguish HERs from cardiac-related artifacts has hindered our understanding of their underlying network (3).

Here, we investigate HERs across vigilance states using intracranial EEG (SEEG). This approach offers unmatched temporal and spatial resolution, enabling a more precise characterization of the neuronal response to heartbeat and its underlying network. We analyzed intracranial recordings from N = 10 epileptic patients undergoing SEEG for presurgical evaluation. HERs were extracted during wakefulness and sleep stages (N2, N3, and REM) by averaging peristimulus epochs time-locked to the R-peak of the heartbeat. Our results reveal that HERs are consistently present across all vigilance states, with the insula emerging as a key neural source of this evoked potential. These findings highlight the critical role of the insula in cardiac signal processing and advance our understanding of brain-heart interactions across different states of consciousness.

1: Candia-Rivera (2022).

2: Mazzola et al (2023).

3: Park & Blanke (2019).

Many automatic processes seem to escape conscious control. A well-known example is Stroop interference, in which the automatic reading response slows participants when they name the color of a word incongruent with its meaning (e.g., blue written in red). Interestingly, highly suggestible individuals can abolish this effect when given a hypnotic suggestion that they can no longer read (the word blindness effect). According to Cold Control Theory (CCT), the ability to suppress reading is not exclusive to highly suggestible people. It claims that anyone can voluntarily control automatic processes without hypnotic suggestions, as long as they are motivated and believe they can. Instead, CCT explains the differences between highly and lowly suggestible individuals as arising from metacognitive skills, where highly hypnotizable lack metacognitive insight into their intention to act in accordance with suggestions.

Here, we tested these claims. We investigated whether a voluntary word blindness strategy can reduce Stroop interference without hypnosis and whether this reduction is independent of suggestibility, as proposed by CCT. Additionally, we examined whether suggestibility can be explained by metacognitive insight into the intention to act. We tested 170 participants, measuring their suggestibility, their ability to reduce Stroop interference using the voluntary word blindness strategy, and their metacognitive efficiency in a visuomotor control task. Consistent with CCT, we show that individuals can slightly reduce Stroop interference voluntarily, regardless of their suggestibility. However, contrary to CCT, metacognitive efficiency did not predict suggestibility. These findings challenge CCT's assumptions and provide new insights into voluntary control over automatic processes.

The question "What is a self-generated action?" has sparked fruitful debates and inspired promising definitions in the field. Notably, Schuur and Haggard propose defining it as the motor consequence of integrating a range of different types of inputs. While we agree that this definition is insightful, we believe it overlooks an important element: the nature of the inputs (exogenous vs. endogenous). Specifically, we suggest that actions relying more on endogenous inputs are more self-generated compared to actions relying more on exogenous inputs.

Our study explores how the quantity and nature of inputs (exogenous vs. endogenous) influence markers of volition in decision-making, using a Random Dot Kinematic (RDK) task. In this pre-registered EEG study, we recruited 39 participants who judged the direction of RDK stimuli across four conditions. We systematically manipulated the quantity of inputs across four levels: Instructed (one possible choice of direction), Semi-Instructed (four possible choices), Free (eight possible choices), and Recommended (eight possible choices with an exogenous recommendation). Simultaneously, we varied the nature of inputs, from Instructed (exogenous decision-making) to Semi-Instructed/Free (endogenous decision-making), and Recommended (a mix of endogenous and exogenous decision-making).

We used a combination of markers, including self-reported volition, temporal binding, and readiness potential, as both subjective and implicit indicators of volition. Our results show that the nature of inputs (exogenous vs. endogenous) had a greater influence on volitional markers than the quantity of inputs. This study provides novel insights into the interplay between input characteristics and volition, offering a refined framework for understanding self-generated actions.

Recent theories propose that consciousness evolved as part of the episodic memory system, suggesting conscious experience might be a product of episodic memory formation. This adds an important perspective to consciousness science, yet the field remains fragmented with competing, largely disconnected theories. In this work, we provide a unified account of three accounts of consciousness: episodic memory, predictive processing, and Integrated Information Theory (IIT). Specifically, we propose that consciousness emerges from the mechanism through which the brain segments continuous sensory data into discrete events and encodes them into episodic memory. This mechanism is driven by points of high Bayesian surprise, which we demonstrate mathematically correspond to peaks in integrated information (φ). This segmentation not only creates our memories but may generate our subjective experience of the present – providing a formal bridge between episodic memory formation, event cognition, and well-tested theories of consciousness. Our theory generates testable predictions: (1) temporal alignment between event boundaries in memory, peaks in φ, and subjective reports of conscious awareness; (2) disruption of hippocampal-cortical integration should affect both memory formation and the continuity of conscious experience; (3) computational models implementing surprise-driven event segmentation should exhibit emergent properties analogous to conscious report. This framework subsumes multiple accounts of consciousness and offers a novel perspective on why conscious experience appears continuous despite evidence for discrete processing: every time our brain detects and encodes a new event in memory, it may be painting our fleeting portrait of the present moment.