Rethinking delusions: A selective review of delusion research through a computational lens

The truth is in there: Belief processes in the human brain

Belief, defined by William James as the mental state or function of cognizing reality, is a core psychological function with strong influence on emotion and behavior. Furthermore, strong and aberrant beliefs about the world and oneself play important roles in mental disorders. The underlying processes of belief have been the matter of a long debate in philosophy and psychology, and modern neuroimaging techniques can provide insight into the underlying neural processes. Here, we conducted a functional magnetic resonance imaging study with N = 30 healthy participants in which we presented statements about facts, politics, religion, conspiracy theories, and superstition. Participants judged whether they considered them as true (belief) or not (disbelief) and reported their certainty in the decision. We found belief-associated activations in bilateral dorsolateral prefrontal cortex, left superior parietal cortex, and left lateral frontopolar cortex. Disbelief-associated activations were found in an anterior temporal cluster extending into the amygdala. We found a larger deactivation for disbelief than belief in the ventromedial prefrontal cortex that was most pronounced during decisions, suggesting a role of the vmPFC in belief-related decision-making. As a category-specific effect, we found disbelief-associated activation in retrosplenial cortex and parahippocampal gyrus for conspiracy theory statements. Exploratory analyses identified networks centered at anterior cingulate cortex for certainty, and dorsomedial prefrontal cortex for uncertainty. The uncertainty effect identifies a neural substrate for Alexander Bain's notion from 1859 of uncertainty as the real opposite of belief. Taken together, our results suggest a two-factor neural process model of belief with falsehood/veracity and uncertainty/certainty factors.

Examining the influence of self-referential thinking on aberrant salience and jumping to conclusions bias in individuals with schizophrenia-spectrum disorders

BACKGROUND AND OBJECTIVES

Cognitive processes such as aberrant salience and the jumping to conclusions (JTC) bias are implicated in the development of delusions. Self-referential thinking is implicated in this process; however, it is unknown how it may interact with aberrant salience and JTC bias in individuals with schizophrenia-spectrum disorders (SSDs). This study examined associations of self-referential thinking with aberrant salience, JTC bias, and delusion severity, and whether self-referential stimuli led to an increase in aberrant salience and JTC bias in SSDs (n = 20) relative to psychiatrically healthy controls (n = 20).

METHODS

To assess aberrant salience and JTC bias, participants were asked to complete both self-referential and neutral versions of the Salience Attribution Test (SAT) and the Beads Task, as well as self-report measures of aberrant salience and JTC bias.

RESULTS

Self-referential task condition interacted with clinical group to predict JTC beads task scores, such that participants with SSDs exhibited greater levels of JTC bias than psychiatrically healthy controls during the neutral task condition, when controlling for levels of motivation, cognitive insight, and functioning. Self-referential thinking was significantly associated with aberrant salience, JTC bias, and delusion severity.

LIMITATIONS

This experiment examined trait-level relationships between variables, so does not provide information about state-level interrelationships and would benefit from replication using more dynamic methods such as ecological momentary assessment.

CONCLUSIONS

These findings highlight the interrelationships between self-referential thinking, JTC bias, aberrant salience, and delusion severity, in individuals with SSDs, and support the interactive role of self-referential thinking in predicting JTC bias.

A Bayesian model of the jumping-to-conclusions bias and its relationship to psychopathology

The mechanisms by which delusion and anxiety affect the tendency to make hasty decisions (Jumping-to-Conclusions bias) remain unclear. This paper proposes a Bayesian computational model that explores the assignment of evidence weights as a potential explanation of the Jumping-to-Conclusions bias using the Beads Task. We also investigate the Beads Task as a repeated measure by varying the key aspects of the paradigm. The Bayesian model estimations from two online studies showed that higher delusional ideation promoted reduced belief updating but the impact of general and social anxiety on evidence weighting was inconsistent. The altered evidence weighting as a result of a psychopathological trait appeared insufficient in contributing to the Jumping-to-Conclusions bias. Variations in Beads Task aspects significantly affected subjective certainty at the point of decisions but not the number of draws to decisions. Repetitions of the Beads Task are feasible if one assesses the Jumping-to-Conclusions bias using number of draws to decisions.

A new predictive coding model for a more comprehensive account of delusions

Attempts to understand psychosis-the experience of profoundly altered perceptions and beliefs-raise questions about how the brain models the world. Standard predictive coding approaches suggest that it does so by minimising mismatches between incoming sensory evidence and predictions. By adjusting predictions, we converge iteratively on a best guess of the nature of the reality. Recent arguments have shown that a modified version of this framework-hybrid predictive coding-provides a better model of how healthy agents make inferences about external reality. We suggest that this more comprehensive model gives us a richer understanding of psychosis compared with standard predictive coding accounts. In this Personal View, we briefly describe the hybrid predictive coding model and show how it offers a more comprehensive account of the phenomenology of delusions, thereby providing a potentially powerful new framework for computational psychiatric approaches to psychosis. We also make suggestions for future work that could be important in formalising this novel perspective.

Introspective inference counteracts perceptual distortion

Introspective agents can recognize the extent to which their internal perceptual experiences deviate from the actual states of the external world. This ability, also known as insight, is critically required for reality testing and is impaired in psychosis, yet little is known about its cognitive underpinnings. We develop a Bayesian modeling framework and a psychophysics paradigm to quantitatively characterize this type of insight while people experience a motion after-effect illusion. People can incorporate knowledge about the illusion into their decisions when judging the actual direction of a motion stimulus, compensating for the illusion (and often overcompensating). Furthermore, confidence, reaction-time, and pupil-dilation data all show signatures consistent with inferential adjustments in the Bayesian insight model. Our results suggest that people can question the veracity of what they see by making insightful inferences that incorporate introspective knowledge about internal distortions.

Bayesian accounts of perceptual decisions in the nonclinical continuum of psychosis: Greater imprecision in both top-down and bottom-up processes

Neurocomputational accounts of psychosis propose mechanisms for how information is integrated into a predictive model of the world, in attempts to understand the occurrence of altered perceptual experiences. Conflicting Bayesian theories postulate aberrations in either top-down or bottom-up processing. The top-down theory predicts an overreliance on prior beliefs or expectations resulting in aberrant perceptual experiences, whereas the bottom-up theory predicts an overreliance on current sensory information, as aberrant salience is directed towards objectively uninformative stimuli. This study empirically adjudicates between these models. We use a perceptual decision-making task in a neurotypical population with varying degrees of psychotic-like experiences. Bayesian modelling was used to compute individuals' reliance on prior relative to sensory information. Across two datasets (discovery dataset n = 363; independent replication in validation dataset n = 782) we showed that psychotic-like experiences were associated with an overweighting of sensory information relative to prior expectations, which seem to be driven by decreased precision afforded to prior information. However, when prior information was more uncertain, participants with greater psychotic-like experiences encoded sensory information with greater noise. Greater psychotic-like experiences were associated with aberrant precision in the encoding both prior and likelihood information, which we suggest may be related to generally heightened perceptions of task instability. Our study lends empirical support to notions of both weaker bottom-up and weaker (rather than stronger) top-down perceptual processes, as well as aberrancies in belief updating that extend into the non-clinical continuum of psychosis.

Editorial: What is computational psychopathology, and why do we need it?

Computational Psychopathology is an emerging discipline, which is based around the theoretical and mechanistic focus of explanatory psychopathology and computational psychiatry, and reflects the general move in psychiatric research away from whole disorders to component symptoms or transdiagnostic processes. In this Editorial, we present a brief summary of these disciplines and how they combine together to form a 'Computational Psychopathology', and present a brief possible taxonomy. We highlight the papers that form part of this Special Issue, along with their place in our putative taxonomy. We conclude this Editorial by highlighting the benefits of a Computational Psychopathology for research into mental health.

Relationships between cognitive biases, decision-making, and delusions

Multiple measures of decision-making under uncertainty (e.g. jumping to conclusions (JTC), bias against disconfirmatory evidence (BADE), win-switch behavior, random exploration) have been associated with delusional thinking in independent studies. Yet, it is unknown whether these variables explain shared or unique variance in delusional thinking, and whether these relationships are specific to paranoia or delusional ideation more broadly. Additionally, the underlying computational mechanisms require further investigation. To investigate these questions, task and self-report data were collected in 88 individuals (46 healthy controls, 42 schizophrenia-spectrum) and included measures of cognitive biases and behavior on probabilistic reversal learning and explore/exploit tasks. Of those, only win-switch rate significantly differed between groups. In regression, reversal learning performance, random exploration, and poor evidence integration during BADE showed significant, independent associations with paranoia. Only self-reported JTC was associated with delusional ideation, controlling for paranoia. Computational parameters increased the proportion of variance explained in paranoia. Overall, decision-making influenced by strong volatility and variability is specifically associated with paranoia, whereas self-reported hasty decision-making is specifically associated with other themes of delusional ideation. These aspects of decision-making under uncertainty may therefore represent distinct cognitive processes that, together, have the potential to worsen delusional thinking across the psychosis spectrum.

Computational mechanisms of belief updating in relation to psychotic-like experiences

Introduction

Psychotic-like experiences (PLEs) may occur due to changes in weighting prior beliefs and new evidence in the belief updating process. It is still unclear whether the acquisition or integration of stable beliefs is altered, and whether such alteration depends on the level of environmental and belief precision, which reflects the associated uncertainty. This motivated us to investigate uncertainty-related dynamics of belief updating in relation to PLEs using an online study design.

Methods

We selected a sample (n = 300) of participants who performed a belief updating task with sudden change points and provided self-report questionnaires for PLEs. The task required participants to observe bags dropping from a hidden helicopter, infer its position, and dynamically update their belief about the helicopter's position. Participants could optimize performance by adjusting learning rates according to inferred belief uncertainty (inverse prior precision) and the probability of environmental change points. We used a normative learning model to examine the relationship between adherence to specific model parameters and PLEs.

Results

PLEs were linked to lower accuracy in tracking the outcome (helicopter location) (β = 0.26 ± 0.11, p = 0.018) and to a smaller increase of belief precision across observations after a change point (β = -0.003 ± 0.0007, p < 0.001). PLEs were related to slower belief updating when participants encountered large prediction errors (β = -0.03 ± 0.009, p = 0.001). Computational modeling suggested that PLEs were associated with reduced overall belief updating in response to prediction errors (βPE = -1.00 ± 0.45, p = 0.028) and reduced modulation of updating at inferred environmental change points (βCPP = -0.84 ± 0.38, p = 0.023).

Discussion

We conclude that PLEs are associated with altered dynamics of belief updating. These findings support the idea that the process of balancing prior belief and new evidence, as a function of environmental uncertainty, is altered in PLEs, which may contribute to the development of delusions. Specifically, slower learning after large prediction errors in people with high PLEs may result in rigid beliefs. Disregarding environmental change points may limit the flexibility to establish new beliefs in the face of contradictory evidence. The present study fosters a deeper understanding of inferential belief updating mechanisms underlying PLEs.

Formalising social representation to explain psychiatric symptoms

Recent work in social cognition has moved beyond a focus on how people process social rewards to examine how healthy people represent other agents and how this is altered in psychiatric disorders. However, formal modelling of social representation has not kept pace with these changes, impeding our understanding of how core aspects of social cognition function, and fail, in psychopathology. Here, we suggest that belief-based computational models provide a basis for an integrated sociocognitive approach to psychiatry, with the potential to address important but unexamined pathologies of social representation, such as maladaptive schemas and illusory social agents.

Bayes' Theorem in Neurocritical Care: Principles and Practice

Patients with critical neurological illness are diverse. As a result of the heterogeneity of this patient population, standardized approaches to patient management might not confer benefit. A precision medicine approach to neurocritical care is therefore urgently needed to improve our understanding of neurocritical illness and the care provided to this vulnerable cohort. Research designs and approaches based on Bayesian models have the potential to meet this need, as they are specifically designed to evolve with emerging evidence. This adaptability provides a benefit over the popular frequentist statistical approach, as it provides a way of adjusting hypotheses and trial procedures to maximize efficacy. This review summarizes the current state of knowledge on Bayes' theorem, and its potential applications to the field of neurocritical care. We review the basic principles underlying Bayes' theorem, compare the use of Bayesian versus frequentist statistics in medicine, and discuss the relevance of Bayesian statistics to the field of neuroscience and to clinical research. Finally, we explore the potential benefits of employing Bayesian methods within the field of neurocritical care as a steppingstone toward implementing precision medicine approaches to improve patient outcomes for complex, heterogeneous disorders.

Everything is connected: Inference and attractors in delusions

Delusions are, by popular definition, false beliefs that are held with certainty and resistant to contradictory evidence. They seem at odds with the notion that the brain at least approximates Bayesian inference. This is especially the case in schizophrenia, a disorder thought to relate to decreased - rather than increased - certainty in the brain's model of the world. We use an active inference Markov decision process model (a Bayes-optimal decision-making agent) to perform a simple task involving social and non-social inferences. We show that even moderate changes in some model parameters - decreasing confidence in sensory input and increasing confidence in states implied by its own (especially habitual) actions - can lead to delusions as defined above. Incorporating affect in the model increases delusions, specifically in the social domain. The model also reproduces some classic psychological effects, including choice-induced preference change, and an optimism bias in inferences about oneself. A key observation is that no change in a single parameter is both necessary and sufficient for delusions; rather, delusions arise due to conditional dependencies that create 'basins of attraction' which trap Bayesian beliefs. Simulating the effects of antidopaminergic antipsychotics - by reducing the model's confidence in its actions - demonstrates that the model can escape from these attractors, through this synthetic pharmacotherapy.

Postdiction in Visual Awareness in Schizophrenia

Background: The mistiming of predictive thought and real perception leads to postdiction in awareness. Individuals with high delusive thinking confuse prediction and perception, which results in impaired reality testing. The present observational study investigated how antipsychotic medications and cognitive-behavioral therapy (CBT) modulate postdiction in schizophrenia. We hypothesized that treatment reduces postdiction, especially when antipsychotics and CBT are combined. Methods: We enrolled patients with schizophrenia treated in a natural clinical setting and not in a randomized controlled trial. We followed up two schizophrenia groups matched for age, sex, education, and illness duration: patients on antipsychotics (n = 25) or antipsychotics plus CBT (n = 25). The treating clinician assigned the patients to the two groups. Participants completed a postdiction and a temporal discrimination task at weeks 0 and 12. Results: At week 0, postdiction was enhanced in patients relative to controls at a short prediction–perception time interval, which correlated with PANSS positive symptoms and delusional conviction. At week 12, postdiction was reduced in schizophrenia, especially when they received antipsychotics plus CBT. Patients with schizophrenia were also impaired on the temporal discrimination task, which did not change during the treatment. During the 12-week observational period, all PANSS scores were significantly reduced in both clinical groups, but the positive symptoms and emotional distress exhibited a more pronounced response in the antipsychotics plus CBT group. Conclusion: Perceptual postdiction is a putative neurocognitive marker of delusive thinking. Combined treatment with antipsychotics and CBT significantly ameliorates abnormally elevated postdiction in schizophrenia.

Unstable Belief Formation and Slowed Decision-making: Evidence That the Jumping-to-Conclusions Bias in Schizophrenia Is Not Linked to Impulsive Decision-making

BACKGROUND

Jumping-to-conclusions (JTC) is a prominent reasoning bias in schizophrenia (SCZ). While it has been linked to not only psychopathological abnormalities (delusions and impulsive decision-making) but also unstable belief formation, its origin remains unclear. We here directly test to which extend JTC is associated with delusional ideation, impulsive decision-making, and unstable belief formation.

METHODS

In total, 45 SCZ patients were compared with matched samples of 45 patients with major depressive disorder (MDD) and 45 healthy controls (HC) as delusions and JTC also occur in other mental disorders and the general population. Participants performed a probabilistic beads task. To test the association of JTC with measures of delusions (Positive and Negative Syndrome Scale [PANSS]positive, PANSSpositive-factor, and Peter Delusions Inventory [PDI]), Bayesian linear regressions were computed. For the link between JTC and impulsive decision-making and unstable beliefs, we conducted between-group comparisons of "draws to decision" (DTD), "decision times" (DT), and "disconfirmatory evidence scores" (DES).

RESULTS

Bayesian regression obtained no robust relationship between PDI and DTD (all |R2adj| ≤ .057, all P ≥ .022, all Bayes Factors [BF01] ≤ 0.046; α adj = .00833). Compared with MDD and HC, patients with SCZ needed more time to decide (significantly higher DT in ambiguous trials: all P ≤ .005, r2 ≥ .216; numerically higher DT in other trials). Further, SCZ had unstable beliefs about the correct source jar whenever unexpected changes in bead sequences (disconfirmatory evidence) occurred (compared with MDD: all P ≤ .004 and all r2 ≥ .232; compared with HC: numerically higher DES). No significant correlation was observed between DT and DTD (all P ≥ .050).

CONCLUSIONS

Our findings point toward a relationship of JTC with unstable belief formation and do not support the assumption that JTC is associated with impulsive decision-making.

Biases and Variability from Costly Bayesian Inference

When humans infer underlying probabilities from stochastic observations, they exhibit biases and variability that cannot be explained on the basis of sound, Bayesian manipulations of probability. This is especially salient when beliefs are updated as a function of sequential observations. We introduce a theoretical framework in which biases and variability emerge from a trade-off between Bayesian inference and the cognitive cost of carrying out probabilistic computations. We consider two forms of the cost: a precision cost and an unpredictability cost; these penalize beliefs that are less entropic and less deterministic, respectively. We apply our framework to the case of a Bernoulli variable: the bias of a coin is inferred from a sequence of coin flips. Theoretical predictions are qualitatively different depending on the form of the cost. A precision cost induces overestimation of small probabilities, on average, and a limited memory of past observations, and, consequently, a fluctuating bias. An unpredictability cost induces underestimation of small probabilities and a fixed bias that remains appreciable even for nearly unbiased observations. The case of a fair (equiprobable) coin, however, is singular, with non-trivial and slow fluctuations in the inferred bias. The proposed framework of costly Bayesian inference illustrates the richness of a 'resource-rational' (or 'bounded-rational') picture of seemingly irrational human cognition.