Encoding Of Prediction Errors Research Articles

The Role of the Insula in Chronic Pain and Associated Structural Changes: An Integrative Review.

Chronic pain affects a substantial portion of the global population, significantly impacting quality of life and well-being. This condition involves complex mechanisms, including dysfunction of the autonomic nervous system, which plays a crucial role in pain perception.The insula, a key brain region involved in pain processing, plays a critical role in pain perception and modulation. Lesions in the insula can result in pain asymbolia, where pain perception remains intact but emotional responses are inappropriate. The insula is anatomically and functionally divided into anterior and posterior regions, with the posterior insula processing nociceptive input based on intensity and location before relaying it to the anterior insula for emotional mediation. Understanding the insula's intricate role in pain processing is crucial, as it is involved in encoding prediction errors and mediating emotional dimensions of pain perception.The focus of this review wason synthesizing existing literature on the role of the insula in chronic pain and associated structural changes. The goal was to integrate findings from various sources to provide a comprehensive overview of the topic.The search strategy included a combination of Medical Subject Headings (MeSH) and relevant keywords related to insulaand chronic pain. The following databases were surveyed: PubMed, Embase, Scopus, and Web of Science. We identified a total of 2515 articles, and after following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline based on eligibility criteria, 46 articles were used to synthesize this review. Our study highlights the pivotal role of the insula in chronic pain processing and associated structural changes, integrating findings from diverse studies and neuroimaging investigations. Beyond mere pain sensation, the insula contributes to emotional awareness, attention, and salience detection within the pain network.Various chronic pain conditions reveal alterations in insular activity and connectivity, accompanied by changes in gray matter volume and neurochemical profiles. Interventions targeting the insula show promise in alleviating chronic pain symptoms. However, further research is needed to understand underlying mechanisms, which can aid in developing more effective therapeutic interventions for pain.

The parietal cortex has a causal role in ambiguity computations in humans.

Humans often face the challenge of making decisions between ambiguous options. The level of ambiguity in decision-making has been linked to activity in the parietal cortex, but its exact computational role remains elusive. To test the hypothesis that the parietal cortex plays a causal role in computing ambiguous probabilities, we conducted consecutive fMRI and TMS-EEG studies. We found that participants assigned unknown probabilities to objective probabilities, elevating the uncertainty of their decisions. Parietal cortex activity correlated with the objective degree of ambiguity and with a process that underestimates the uncertainty during decision-making. Conversely, the midcingulate cortex (MCC) encodes prediction errors and increases its connectivity with the parietal cortex during outcome processing. Disruption of the parietal activity increased the uncertainty evaluation of the options, decreasing cingulate cortex oscillations during outcome evaluation and lateral frontal oscillations related to value ambiguous probability. These results provide evidence for a causal role of the parietal cortex in computing uncertainty during ambiguous decisions made by humans.

Multiple Concurrent Predictions Inform Prediction Error in the Human Auditory Pathway.

The key assumption of the predictive coding framework is that internal representations are used to generate predictions on how the sensory input will look like in the immediate future. These predictions are tested against the actual input by the so-called prediction error units, which encode the residuals of the predictions. What happens to prediction errors, however, if predictions drawn by different stages of the sensory hierarchy contradict each other? To answer this question, we conducted two fMRI experiments while female and male human participants listened to sequences of sounds: pure tones in the first experiment and frequency-modulated sweeps in the second experiment. In both experiments, we used repetition to induce predictions based on stimulus statistics (stats-informed predictions) and abstract rules disclosed in the task instructions to induce an orthogonal set of (task-informed) predictions. We tested three alternative scenarios: neural responses in the auditory sensory pathway encode prediction error with respect to (1) the stats-informed predictions, (2) the task-informed predictions, or (3) a combination of both. Results showed that neural populations in all recorded regions (bilateral inferior colliculus, medial geniculate body, and primary and secondary auditory cortices) encode prediction error with respect to a combination of the two orthogonal sets of predictions. The findings suggest that predictive coding exploits the non-linear architecture of the auditory pathway for the transmission of predictions. Such non-linear transmission of predictions might be crucial for the predictive coding of complex auditory signals like speech.Significance Statement Sensory systems exploit our subjective expectations to make sense of an overwhelming influx of sensory signals. It is still unclear how expectations at each stage of the processing pipeline are used to predict the representations at the other stages. The current view is that this transmission is hierarchical and linear. Here we measured fMRI responses in auditory cortex, sensory thalamus, and midbrain while we induced two sets of mutually inconsistent expectations on the sensory input, each putatively encoded at a different stage. We show that responses at all stages are concurrently shaped by both sets of expectations. The results challenge the hypothesis that expectations are transmitted linearly and provide for a normative explanation of the non-linear physiology of the corticofugal sensory system.

Reward is not reward: Differential impacts of primary and secondary rewards on expectation, outcome, and prediction error in the human brain's reward processing regions

According to their nature, rewarding stimuli are classified as primary (e.g., food, sex) and secondary (e.g., money) rewards. Neuroimaging studies have provided valuable insights in neural reward processing and its various aspects including reward expectation, outcome and prediction error encoding. However, there is only limited evidence of whether the two different types of rewards are processed in common or distinct brain areas, in particular when considering the different functions of reward processing. We analyzed a sample of 42 healthy, male participants using task-based functional magnetic resonance imaging (fMRI) during a variant of the monetary incentive delay task. We aimed to investigate the effects of three different rewarding stimuli—two primary (food and sex) and one secondary (money)—on the various functions of reward processing. To provide a thorough description, we focused on 12 brain regions of interest and utilized the Bayes factor bound (BFB) to express stimulus-related main effects and pairwise differences at different levels of evidence, ranging from weak to decisive. Our results revealed a dominance of sexually charged stimuli in engaging the brain's reward structures for all investigated aspects of reward processing. Nevertheless, the ventral tegmental area, amygdala, ventral caudate, ventromedial prefrontal cortex, subgenual anterior cingulate cortex, and lateral orbitofrontal cortex were activated by both primary and secondary reward outcomes. For other reward processing functions, i.e., reward expectation and the prediction error, effects of the different stimuli were weaker, and effects from one reward type cannot easily be generalized to the other.

Distinct roles of delta- and theta-band neural tracking for sharpening and predictive coding of multi-level speech features during spoken language processing.

The brain tracks and encodes multi-level speech features during spoken language processing. It is evident that this speech tracking is dominant at low frequencies (<8 Hz) including delta and theta bands. Recent research has demonstrated distinctions between delta- and theta-band tracking but has not elucidated how they differentially encode speech across linguistic levels. Here, we hypothesised that delta-band tracking encodes prediction errors (enhanced processing of unexpected features) while theta-band tracking encodes neural sharpening (enhanced processing of expected features) when people perceive speech with different linguistic contents. EEG responses were recorded when normal-hearing participants attended to continuous auditory stimuli that contained different phonological/morphological and semantic contents: (1) real-words, (2) pseudo-words and (3) time-reversed speech. We employed multivariate temporal response functions to measure EEG reconstruction accuracies in response to acoustic (spectrogram), phonetic and phonemic features with the partialling procedure that singles out unique contributions of individual features. We found higher delta-band accuracies for pseudo-words than real-words and time-reversed speech, especially during encoding of phonetic features. Notably, individual time-lag analyses showed that significantly higher accuracies for pseudo-words than real-words started at early processing stages for phonetic encoding (<100 ms post-feature) and later stages for acoustic and phonemic encoding (>200 and 400 ms post-feature, respectively). Theta-band accuracies, on the other hand, were higher when stimuli had richer linguistic content (real-words > pseudo-words > time-reversed speech). Such effects also started at early stages (<100 ms post-feature) during encoding of all individual features or when all features were combined. We argue these results indicate that delta-band tracking may play a role in predictive coding leading to greater tracking of pseudo-words due to the presence of unexpected/unpredicted semantic information, while theta-band tracking encodes sharpened signals caused by more expected phonological/morphological and semantic contents. Early presence of these effects reflects rapid computations of sharpening and prediction errors. Moreover, by measuring changes in EEG alpha power, we did not find evidence that the observed effects can be solitarily explained by attentional demands or listening efforts. Finally, we used directed information analyses to illustrate feedforward and feedback information transfers between prediction errors and sharpening across linguistic levels, showcasing how our results fit with the hierarchical Predictive Coding framework. Together, we suggest the distinct roles of delta and theta neural tracking for sharpening and predictive coding of multi-level speech features during spoken language processing.

Stimulus-Specific Prediction Error Neurons in Mouse Auditory Cortex.

Comparing expectation with experience is an important neural computation performed throughout the brain and is a hallmark of predictive processing. Experiments that alter the sensory outcome of an animal's behavior reveal enhanced neural responses to unexpected self-generated stimuli, indicating that populations of neurons in sensory cortex may reflect prediction errors (PEs), mismatches between expectation and experience. However, enhanced neural responses to self-generated stimuli could also arise through nonpredictive mechanisms, such as the movement-based facilitation of a neuron's inherent sound responses. If sensory prediction error neurons exist in sensory cortex, it is unknown whether they manifest as general error responses, or respond with specificity to errors in distinct stimulus dimensions. To answer these questions, we trained mice of either sex to expect the outcome of a simple sound-generating behavior and recorded auditory cortex activity as mice heard either the expected sound or sounds that deviated from expectation in one of multiple distinct dimensions. Our data reveal that the auditory cortex learns to suppress responses to self-generated sounds along multiple acoustic dimensions simultaneously. We identify a distinct population of auditory cortex neurons that are not responsive to passive sounds or to the expected sound but that encode prediction errors. These prediction error neurons are abundant only in animals with a learned motor-sensory expectation, and encode one or two specific violations rather than a generic error signal. Together, these findings reveal that cortical predictions about self-generated sounds have specificity in multiple simultaneous dimensions and that cortical prediction error neurons encode specific violations from expectation.SIGNIFICANCE STATEMENT Audette et. al record neural activity in the auditory cortex while mice perform a sound-generating forelimb movement and measure neural responses to sounds that violate an animal's expectation in different ways. They find that predictions about self-generated sounds are highly specific across multiple stimulus dimensions and that a population of typically nonsound-responsive neurons respond to sounds that violate an animal's expectation in a specific way. These results identify specific prediction error (PE) signals in the mouse auditory cortex and suggest that errors may be calculated early in sensory processing.

Omission responses in local field potentials in rat auditory cortex

BackgroundNon-invasive recordings of gross neural activity in humans often show responses to omitted stimuli in steady trains of identical stimuli. This has been taken as evidence for the neural coding of prediction or prediction error. However, evidence for such omission responses from invasive recordings of cellular-scale responses in animal models is scarce. Here, we sought to characterise omission responses using extracellular recordings in the auditory cortex of anaesthetised rats. We profiled omission responses across local field potentials (LFP), analogue multiunit activity (AMUA), and single/multi-unit spiking activity, using stimuli that were fixed-rate trains of acoustic noise bursts where 5% of bursts were randomly omitted.ResultsSignificant omission responses were observed in LFP and AMUA signals, but not in spiking activity. These omission responses had a lower amplitude and longer latency than burst-evoked sensory responses, and omission response amplitude increased as a function of the number of preceding bursts.ConclusionsTogether, our findings show that omission responses are most robustly observed in LFP and AMUA signals (relative to spiking activity). This has implications for models of cortical processing that require many neurons to encode prediction errors in their spike output.

Recurrent predictive coding models for associative memory employing covariance learning.

The computational principles adopted by the hippocampus in associative memory (AM) tasks have been one of the most studied topics in computational and theoretical neuroscience. Recent theories suggested that AM and the predictive activities of the hippocampus could be described within a unitary account, and that predictive coding underlies the computations supporting AM in the hippocampus. Following this theory, a computational model based on classical hierarchical predictive networks was proposed and was shown to perform well in various AM tasks. However, this fully hierarchical model did not incorporate recurrent connections, an architectural component of the CA3 region of the hippocampus that is crucial for AM. This makes the structure of the model inconsistent with the known connectivity of CA3 and classical recurrent models such as Hopfield Networks, which learn the covariance of inputs through their recurrent connections to perform AM. Earlier PC models that learn the covariance information of inputs explicitly via recurrent connections seem to be a solution to these issues. Here, we show that although these models can perform AM, they do it in an implausible and numerically unstable way. Instead, we propose alternatives to these earlier covariance-learning predictive coding networks, which learn the covariance information implicitly and plausibly, and can use dendritic structures to encode prediction errors. We show analytically that our proposed models are perfectly equivalent to the earlier predictive coding model learning covariance explicitly, and encounter no numerical issues when performing AM tasks in practice. We further show that our models can be combined with hierarchical predictive coding networks to model the hippocampo-neocortical interactions. Our models provide a biologically plausible approach to modelling the hippocampal network, pointing to a potential computational mechanism during hippocampal memory formation and recall, which employs both predictive coding and covariance learning based on the recurrent network structure of the hippocampus.

Predictive encoding of pure tones and FM-sweeps in the human auditory cortex.

Expectations substantially influence perception, but the neural mechanisms underlying this influence are not fully understood. A prominent view is that sensory neurons encode prediction error with respect to expectations on upcoming sensory input. Although the encoding of prediction error has been previously demonstrated in the human auditory cortex (AC), previous studies often induced expectations using stimulus repetition, potentially confounding prediction error with neural habituation. These studies also measured AC as a single population, failing to consider possible predictive specializations of different AC fields. Moreover, the few studies that considered prediction error to stimuli other than pure tones yielded conflicting results. Here, we used functional magnetic resonance imaging (fMRI) to systematically investigate prediction error to subjective expectations in auditory cortical fields Te1.0, Te1.1, Te1.2, and Te3, and two types of stimuli: pure tones and frequency modulated (FM) sweeps. Our results show that prediction error is elicited with respect to the participants' expectations independently of stimulus repetition and similarly expressed across auditory fields. Moreover, despite the radically different strategies underlying the decoding of pure tones and FM-sweeps, both stimulus modalities were encoded as prediction error in most fields of AC. Altogether, our results provide unequivocal evidence that predictive coding is the general encoding mechanism in AC.

Oxytocin modulates neurocomputational mechanisms underlying prosocial reinforcement learning.

Humans often act in the best interests of others. However, how we learn which actions result in good outcomes for other people and the neurochemical systems that support this 'prosocial learning' remain poorly understood. Using computational models of reinforcement learning, functional magnetic resonance imaging and dynamic causal modelling, we examined how different doses of intranasal oxytocin, a neuropeptide linked to social cognition, impact how people learn to benefit others (prosocial learning) and whether this influence could be dissociated from how we learn to benefit ourselves (self-oriented learning). We show that a low dose of oxytocin prevented decreases in prosocial performance over time, despite no impact on self-oriented learning. Critically, oxytocin produced dose-dependent changes in the encoding of prediction errors (PE) in the midbrain-subgenual anterior cingulate cortex (sgACC) pathway specifically during prosocial learning. Our findings reveal a new role of oxytocin in prosocial learning by modulating computations of PEs in the midbrain-sgACC pathway.

P272. Diminished ‘Punishment Learning’ in Patients with Treatment-Resistant Depression

Human neuroimaging studies demonstrate diminished reward learning in dopaminergic regions known to encode prediction errors between expected and experienced rewards in patients with depression. However, differences in punishment learning and how these may vary in patients responsive or not to antidepressants is unclear. We utilized a novel valence-partitioned reinforcement learning model to examine cohort-specific differences in punishment learning.

Adaptive encoding based lossless data hiding method for VQ compressed images using tabu search

A novel joint image coding and reversible data hiding method for vector quantization (VQ) compressed images is proposed in this paper. Since the original VQ indices often exhibit uncorrelated, a rearrangement of them by considering their correlations may benefit the prediction performance so as to reduce the required bitrate. In this study, the tabu search algorithm is employed to rearrange codewords by fully exploiting their neighboring correlations, yielding a moepressible rearranged indices. By combining more highly-correlated indices of a to-be-predicted index into prediction, the improved linear regression method is then applied to achieve a sharper prediction-error histogram and less required additional information. After prediction, an adaptive run-length encoding method is presented to encode prediction errors, thereby eliminating unnecessary indicators. Experimental results demonstrate that the proposed method effectively reduces the bitrate of the compressed image while providing a comparable hiding capacity.

Abnormal prediction error processing in schizophrenia and depression

To make adaptive decisions under uncertainty, individuals need to actively monitor the discrepancy between expected outcomes and actual outcomes, known as prediction errors. Reward‐based learning deficits have been shown in both depression and schizophrenia patients. For this study, we compiled studies that investigated prediction error processing in depression and schizophrenia patients and performed a series of meta‐analyses. In both groups, positive t‐maps of prediction error tend to yield striatum activity across studies. The analysis of negative t‐maps of prediction error revealed two large clusters within the right superior and inferior frontal lobes in schizophrenia and the medial prefrontal cortex and bilateral insula in depression. The concordant posterior cingulate activity was observed in both patient groups, more prominent in the depression group and absent in the healthy control group. These findings suggest a possible role in dopamine‐rich areas associated with the encoding of prediction errors in depression and schizophrenia.

Dissociable dopamine dynamics for learning and motivation.

The dopamine projection from ventral tegmental area (VTA) to nucleus accumbens (NAc) is critical for motivation to work for rewards, and reward-driven learning. How dopamine supports both functions is unclear. Dopamine spiking can encode prediction errors, vital learning signals in computational theories of adaptive behavior. By contrast, dopamine release ramps up as animals approach rewards, mirroring reward expectation. This mismatch might reflect differences in behavioral tasks, slower changes in dopamine cell spiking, or spike-independent modulation of dopamine release. Here we compare spiking of identified VTA dopamine cells with NAc dopamine release in the same decision-making task. Cues indicating upcoming reward increased both spiking and release. Yet NAc core dopamine release also covaried with dynamically-evolving reward expectations, without corresponding changes in VTA dopamine cell spiking. Our results suggest a fundamental difference in how dopamine release is regulated to achieve distinct functions: broadcast burst signals promote learning, while local control drives motivation.

Neurocomputational mechanisms of adaptive learning in social exchanges.

Prior work on prosocial and self-serving behavior in human economic exchanges has shown that counterparts' high social reputations bias striatal reward signals and elicit cooperation, even when such cooperation is disadvantageous. This phenomenon suggests that the human striatum is modulated by the other's social value, which is insensitive to the individual's own choices to cooperate or defect. We tested an alternative hypothesis that, when people learn from their interactions with others, they encode prediction error updates with respect to their own policy. Under this policy update account striatal signals would reflect positive prediction errors when the individual's choices correctly anticipated not only the counterpart's cooperation but also defection. We examined behavior in three samples using reinforcement learning and model-free analyses and performed an fMRI study of striatal learning signals. In order to uncover the dynamics of goal-directed learning, we introduced reversals in the counterpart's behavior and provided counterfactual (would-be) feedback when the individual chose not to engage with the counterpart. Behavioral data and model-derived prediction error maps (in both whole-brain and a priori striatal region of interest analyses) supported the policy update model. Thus, as people continually adjust their rate of cooperation based on experience, their behavior and striatal learning signals reveal a self-centered instrumental process corresponding to reciprocal altruism.

Feature-specific prediction errors and surprise across macaque fronto-striatal circuits

To adjust expectations efficiently, prediction errors need to be associated with the precise features that gave rise to the unexpected outcome, but this credit assignment may be problematic if stimuli differ on multiple dimensions and it is ambiguous which feature dimension caused the outcome. Here, we report a potential solution: neurons in four recorded areas of the anterior fronto-striatal networks encode prediction errors that are specific to feature values of different dimensions of attended multidimensional stimuli. The most ubiquitous prediction error occurred for the reward-relevant dimension. Feature-specific prediction error signals a) emerge on average shortly after non-specific prediction error signals, b) arise earliest in the anterior cingulate cortex and later in dorsolateral prefrontal cortex, caudate and ventral striatum, and c) contribute to feature-based stimulus selection after learning. Thus, a widely-distributed feature-specific eligibility trace may be used to update synaptic weights for improved feature-based attention.

Motivation in the Service of Allostasis: The Role of anterior Mid Cingulate Cortex.

In this article, we suggest that motivation serves to anticipate the energy of the body and meet those needs before they arise, called allostasis. We describe motivation as the output of energy computations that include estimates about future energy/metabolic needs and the value of effort required for potential behaviors (i.e., whether the cost of effort is worthwhile). We bring neuroscience evidence to bear to support this hypothesis. We outline a system of brain networks that have been shown to be important for motivation, and focus in on one hub in this network, the anterior mid-cingulate cortex (aMCC), and discuss its importance for establishing motivation in the service of allostasis. We present evidence that the aMCC, positioned at the intersection of multiple brain networks, is wired to integrate signals relating to allostasis with its sensory consequences, termed interoception, as well as with cognitive control processes, sensory and motor functions. This integration guides the nervous system towards the optimal effort required to achieve a desired goal. Across a variety of task domains, we discuss the role of aMCC in motivation, including a) processing of the value of prior and expected rewards, b) assessment of energetic costs in the brain and the body, c) selectively learning and encoding prediction errors (unexpected changes) that are relevant for allostasis, d) computations for monitoring of internal states of the body and e) modulating the internal state of the body to prepare for action. Finally, we discuss the link between individual differences in aMCC processing and variation in two extreme ends of the range of motivational states, tenacity and apathy.

Differential Roles of the Salience Network During Prediction Error Encoding and Facial Emotion Processing Among Female Adolescent Assault Victims

BackgroundEarly-life assaultive violence exposure is a potent risk factor for posttraumatic stress disorder (PTSD) and other mood and anxiety disorders. Neurocircuitry models posit that increased risk is mediated by heightened emotion processing in a salience network including the dorsal anterior cingulate cortex, anterior insula, and amygdala. However, the processes of reinforcement learning (RL) also engage the salience network and are implicated in responses to early-life trauma and PTSD. To define their relative roles in response to early-life trauma and PTSD symptoms, the current study compared engagement of the salience network during emotion processing and RL as a function of early-life assault exposure. MethodsAdolescent girls (n = 30 girls who had previously been physically or sexually assaulted; n = 30 healthy girls for comparison) 11 to 17 years of age completed two types of tasks during functional magnetic resonance imaging: a facial emotion processing task and an RL task using either social or nonsocial stimuli. Independent component analysis was used to identify a salience network and characterize its engagement in response to emotion processing and prediction error encoding during the RL tasks. ResultsAssault was related to greater reactivity of the salience network during emotion processing. By contrast, we found lesser encoding of negative prediction errors in the salience network, particularly during the social RL task, in girls who had been assaulted. The dysfunction of salience network activity during emotion processing and prediction error encoding was not associated with PTSD symptoms. ConclusionsThese results suggest that hyper- versus hypoactivity of the salience network among trauma-exposed youths depends on the cognitive-affective domain.

Oxytocin attenuates trust as a subset of more general reinforcement learning, with altered reward circuit functional connectivity in males

Oxytocin (OT) is an endogenous neuropeptide that, while originally thought to promote trust, has more recently been found to be context-dependent. Here we extend experimental paradigms previously restricted to de novo decision-to-trust, to a more realistic environment in which social relationships evolve in response to iterative feedback over twenty interactions. In a randomized, double blind, placebo-controlled within-subject/crossover experiment of human adult males, we investigated the effects of a single dose of intranasal OT (40 IU) on Bayesian expectation updating and reinforcement learning within a social context, with associated brain circuit dynamics. Subjects participated in a neuroeconomic task (Iterative Trust Game) designed to probe iterative social learning while their brains were scanned using ultra-high field (7T) fMRI. We modeled each subject's behavior using Bayesian updating of belief-states (“willingness to trust”) as well as canonical measures of reinforcement learning (learning rate, inverse temperature). Behavioral trajectories were then used as regressors within fMRI activation and connectivity analyses to identify corresponding brain network functionality affected by OT. Behaviorally, OT reduced feedback learning, without bias with respect to positive versus negative reward. Neurobiologically, reduced learning under OT was associated with muted communication between three key nodes within the reward circuit: the orbitofrontal cortex, amygdala, and lateral (limbic) habenula. Our data suggest that OT, rather than inspiring feelings of generosity, instead attenuates the brain's encoding of prediction error and therefore its ability to modulate pre-existing beliefs. This effect may underlie OT's putative role in promoting what has typically been reported as ‘unjustified trust’ in the face of information that suggests likely betrayal, while also resolving apparent contradictions with regard to OT's context-dependent behavioral effects.

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

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.