A predictive coding account of bistable perception - a model-based fMRI study

Abstract

In bistable vision, subjective perception wavers between two interpretations of a constant ambiguous stimulus. This dissociation between conscious perception and sensory stimulation has motivated various empirical studies on the neural correlates of bistable perception, but the neurocomputational mechanism behind endogenous perceptual transitions has remained elusive. Here, we recurred to a generic Bayesian framework of predictive coding and devised a model that casts endogenous perceptual transitions as a consequence of prediction errors emerging from residual evidence for the suppressed percept. Data simulations revealed close similarities between the model’s predictions and key temporal characteristics of perceptual bistability, indicating that the model was able to reproduce bistable perception. Fitting the predictive coding model to behavioural data from an fMRI-experiment on bistable perception, we found a correlation across participants between the model parameter encoding perceptual stabilization and the behaviourally measured frequency of perceptual transitions, corroborating that the model successfully accounted for participants’ perception. Formal model comparison with established models of bistable perception based on mutual inhibition and adaptation, noise or a combination of adaptation and noise was used for the validation of the predictive coding model against the established models. Most importantly, model-based analyses of the fMRI data revealed that prediction error time-courses derived from the predictive coding model correlated with neural signal time-courses in bilateral inferior frontal gyri and anterior insulae. Voxel-wise model selection indicated a superiority of the predictive coding model over conventional analysis approaches in explaining neural activity in these frontal areas, suggesting that frontal cortex encodes prediction errors that mediate endogenous perceptual transitions in bistable perception. Taken together, our current work provides a theoretical framework that allows for the analysis of behavioural and neural data using a predictive coding perspective on bistable perception. In this, our approach posits a crucial role of prediction error signalling for the resolution of perceptual ambiguities.

Evaluating the neurophysiological evidence for predictive processing as a model of perception

“…117 Conversely, PP provides a satisfying explanation of these copious backward connections, impugning their consignment to mere recurrent modulations in feedforward schema. 183,206 The explanatory power of this framework, exemplified in simulations of PP that readily account for phenomena ranging from V1 neuron response properties 207 to bistable perception 29 to perceptual illusions, 30,31 has led some to argue that everything the brain does can ultimately be explained in terms of prediction error minimization. 9,17,32 Such a bold contention has inevitably drawn skepticism, much of which has centered on the theory's empirical foundation.…”

“…Although significant progress has been made, 4,42,162 the neurophysiological mechanisms responsible for the extraction of regularities and the generation of experience-based priors are not well understood. Although a range of candidate neural regions have been implicated in tracking a variety of sensory regularities and issuing predictions 29,37,40,53,55,75,120,144,158,162 (see Ref. 41 for a recent review), their precise relationships to the modulations observed in sensory processing are not yet established.…”

“…In the primate visual system, signals transmitted forward through hierarchical levels are expressed in theta-band (4-8 Hz) and gamma-band (>30 Hz) activity, while backward signaling is carried out through beta-band activity (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30). 42,170 PP associates this oscillatory segregation between forward and backward processing 171 with the distinct subpopulations housing error and expectation units, suggesting that oscillatory signatures of error and prediction message passing should be dissociable.…”

From symbols to icons: the return of resemblance in the cognitive neuroscience revolution

“…Kaplan and Craver 2011b) at present, pitched largely at Marr's (1982) "algorithmic" level of description with only tentative proposals about how the functional roles it identifies are realised in cortical circuitry (cf. Brodski et al 2015;Gordon et al 2017;Weilnhammer et al 2017). 6 Nevertheless, it nicely illustrates the points we are making, and manifests how a regulative and thoroughly predictive perspective on brain function harmonises with an iconic approach to cognitive representation.…”

Anterior Insula Reflects Inferential Errors in Value-Based Decision-making and Perception

“…With multistable stimuli however, information conveyed by a constant stimulus is sufficiently uncertain to evoke spontaneous alternations in conscious perception (Rubin, 1921;Boring, 1930;Necker, 1832), suggesting the decision process itself is hidden, with only the outcome made available to the agent. Crucially, ambiguous perception does not rely on dominance time as an index of uncertainty (Brascamp et al, 2005), as switches arise in a stochastic manner., Bi-stable figures have recently been studied in predictive-coding frameworks where perceptual transitions (see also : Dayan, 1998;Hohwy et al, 2008) correlated with anterior insulae and inferior frontal gyrii responses (Weilnhammer et al, 2017), regions also associated with economic uncertainty (Platt & Huettel, 2008;Grinband et al, 2006;Brevers et al, 2015;Preuschoff et al, 2008a;Mohr et al, 2010).…”