The Temporal Interactions between Conscious and Unconscious Perceptual Streams – TICUPS

We took a step-by-step approach to dissect the construction of a conscious percept, using state-of-the-art neuroscientific techniques such as EEG time-frequency analyses, psychophysics or fMRI. The specific questions that we aimed to address are listed below:
1. Neural correlates of MIB: What distinguishes the neural representation of a perceptually suppressed stimulus from the conscious representation of the same stimulus? This was addressed using state-of-the-art single-trial time-frequency EEG analyses.
2. Uploading vs. reloading: What distinguishes the conscious re-appearance of a previously suppressed stimulus (“reloading”) from the conscious perception of a novel stimulus (“uploading”)? This was addressed using advanced EEG time-frequency decomposition methods.
3. Neural correlates of reloading: What brain regions encode the conscious re-appearance of a previously suppressed stimulus (“reloading”)? This was addressed using fMRI.
4. Duration and decay of unconscious representations: For how long does a perceptually suppressed stimulus retain its temporal advantage for conscious access (“reloading”) relative to novel stimuli (“uploading”)? This was addressed psychophysically.
5. Cortical stimulation: Can the 100ms lag between conscious and unconscious perceptual streams be uncovered by another stimulation of visual cortex? This was addressed with the psychophysical masking technique.
6. Cycles of consciousness: What determines the relative rapidity of “reloading” vs. “uploading” conscious processes? Is the 100ms delay fixed (with some room for intrinsic noise), or does it vary from trial to trial? If the latter, do these variations follow a cyclic pattern, i.e. do they co-vary with the phase of a specific ongoing oscillation? This was addressed with the single-trial EEG analysis techniques previously developed in our lab.

We demonstrated, using fMRI, that certain regions of occipital and parietal cortex are specifically implicated in reloading a perceptually suppressed stimulus. Furthermore, EEG analysis revealed that this reloading follows the cycles of a brain oscillation at around 7Hz. Thus, the updating of conscious representations follows a periodic neuronal process. These findings are an important step forward in charting the neuronal correlates of consciousness.

Consciousness is one of the major unsolved scientific problems. Though we understand to some extent the anatomy and physiology of the brain, and can characterize the sequence of neuronal activities following the presentation of a given stimulus (say, a yellow ring), we do not know what part of this activity corresponds to the conscious recognition and phenomenal experience of the stimulus, or equivalently, what would differ if the same stimulus was presented without evoking a conscious percept. Research on this question is important not only because it pushes the boundaries of human knowledge, but also because it can have essential medical, societal or technological applications.

Each of the 6 questions above gave rise to a distinct experiment, half of which were lead by the French partner (CerCo, Toulouse), and the other half by the Taiwanese partner (National Taiwan University), all of them in collaboration with the other partner. All of the experiments produced interpretable results, that were for the most part presented in various international conferences (5 conferences), and gave rise or are expected to give rise to publications in specialized scientific journals (1 article published in a peer-reviewed international journal: “Neuroscience of Consciousness”, 5 more articles still in preparation).

Conscious perception feels instantaneous and effortless. Nonetheless, it is the result of intensive and time-consuming neuronal computations. It is thus not surprising that brain activities generally occur a few tenths of a second after the external event that triggers them. Recently, we revealed that, in addition to such processing latencies inherent in sensory systems, the stream of conscious perception lags even further behind, by about 100ms (Wu, Busch, Fabre-Thorpe & VanRullen, 2009). This discovery was made using Motion-Induced Blindness (MIB), a technique that renders a stimulus (such as a yellow ring) invisible to consciousness, even though it is still present on the retina. A specific manipulation (a flash) allowed us to trigger the return of this ring to consciousness; surprisingly, we discovered that the ring stimulus was consistently perceived to reappear ~100ms before the flash that caused this reappearance. In other words, the "reloading" of the previously established neural representation of the ring into consciousness took at least 100ms less than the normal "uploading" of a novel stimulus (the flash). This implies that the stream of consciousness is normally delayed by ~100ms compared to a stream of fully-formed, but still unconscious representations. What is the cause, and more importantly what is the purpose of this additional delay? Does it reflect, for example, a necessary re-entry (or feedback) of signals before conscious access? Or a minimal activation time for consciousness? The purpose of this project is to methodically investigate this 100ms delay and its neuronal correlates using the latest neuroscientific techniques (EEG single-trial time-frequency analysis, TMS, fMRI), in order to understand the precise temporal relations between the unconscious and the conscious perceptual streams. This, in turn, may lead to important breakthroughs in charting the neuronal correlates of consciousness. The project is a collaborative effort between the two main authors of the original discovery, R. VanRullen (based in France) and C-T. Wu who recently opened his own laboratory in Taiwan.