Feedback information sharing in the human brain reflects bistable perception in the absence of report

Abstract

In the search for the neural basis of conscious experience, perception and its cognitive consequences are typically confounded as neural activity is recorded while participants explicitly report what they experience. Here we present a novel way to disentangle perception from report using eye-movement analysis techniques based on convolutional neural networks and neurodynamical analyses based on information theory. We use a visual bistable stimulus that instantiates two well-known properties of conscious perception: integration and differentiation. At any given moment, observers either experience the stimulus as one integrated unitary percept or as two differentiated objects that are clearly distinct from each other. Using electroencephalography, we show that measures of neural information dynamics of directed information integration and differentiation closely follow participants experience when perceptual transitions were reported. We observed increased directed information from anterior to posterior electrodes (front to back) leading up to the moment the stimulus was reported to switch to the integrated percept and higher information differentiation of anterior signals leading up to reporting the differentiated percept. Crucially, neural integration dynamics were closely linked to perception and even observed in a no-report condition when perceptual transitions were inferred from eye-movements. In contrast, the link between perceptual transitions and neural differentiation was observed solely in the active report condition. Our results therefore suggest a differential role of anterior-posterior network communication vs anterior information differentiation during perception and reporting: while front to back directed information is associated with changes in the content of perception, frontal information differentiation reflects cognitive processes that are consequential of perceptual transitions, not perception per se.

Illusory object recognition is either perceptual or cognitive in origin depending on decision confidence

“…In studies employing such noreport paradigms, neural processing is typically compared between experimental conditions in which participants are aware of specific stimuli, but do not have a task to do and thus merely passively perceive these stimuli, versus a condition where active report is required. Contrasting these conditions, usually also including an unaware condition is thought to separate neural signals associated with perceptual experience per se versus neural signals of post-perceptual processes (or task-relevance in the broader sense) [21][22][23][24][25][26]. We found that decision errors made with high confidence were associated with neural representations reflecting the misperceived object (e.g., a house could be decoded while a face was presented).…”

Drawing the experience dynamics of meditation

“…For example, in two recent studies, Canales-Johnson et al . 15,16 dissociated two phenomenological variants of the same stimulus during an auditory and a visual bistable perception task using frontoparietal information dynamics. In this experimental design the participants were asked to discriminate between two pre-defined, phenomenologically distinct experience phenotypes known to be evoked by the ambiguous stimulus.…”

“…These descriptions were clustered into phenomenologically distinct levels of "preparedness" and explained a significant portion of observed variance in task performance and EEG brain activity.Methods with a wider application potential may require the "front-loading" of phenomenology into the experimental design 5 . For example, in two recent studies, Canales-Johnson et al 15,16 dissociated two phenomenological variants of the same stimulus during an auditory and a visual bistable perception task using frontoparietal information dynamics. In this experimental design the participants were asked to discriminate between two pre-defined, phenomenologically distinct experience phenotypes known to be evoked by the ambiguous stimulus.…”

The perceptual nature of illusory object recognition

“…In the sensory localizer the classifiers' sensitivity was thus tuned mainly to sensory features of the two stimulus categories due to the absence of attention/task-relevance of the face/house images (26,28,29). Therefore, the sensory localizer is reminiscent of a so-called no-report paradigm (30,31), but then used to train localizers instead of being used as the main task of interest (22)(23)(24)32). The decision localizer was, besides sensitive to sensory features, also sensitive to post-perceptual decision processes.…”

“…In studies employing such no-report paradigms, neural processing is typically compared between experimental conditions in which participants are aware of specific stimuli, but do not have a task to do and thus merely passively perceive these stimuli, versus a condition where active report is required. Contrasting these conditions, usually also including an unaware condition is thought to separate neural signals associated with perceptual experience per se versus neural signals of postperceptual processes (or task-relevance in the broader sense) (21)(22)(23)(24)(25)(26). We found that decision errors made with high confidence were associated with neural representations reflecting the misperceived object (e.g., a house could be decoded while a face was presented).…”