Subjective estimates of uncertainty during gambling and impulsivity after subthalamic deep brain stimulation for Parkinson's disease

Computational processes of simultaneous learning of stochasticity and volatility in humans

Making adaptive decisions requires predicting outcomes, and this in turn requires adapting to uncertain environments. This study explores computational challenges in distinguishing two types of noise influencing predictions: volatility and stochasticity. Volatility refers to diffusion noise in latent causes, requiring a higher learning rate, while stochasticity introduces moment-to-moment observation noise and reduces learning rate. Dissociating these effects is challenging as both increase the variance of observations. Previous research examined these factors mostly separately, but it remains unclear whether and how humans dissociate them when they are played off against one another. In two large-scale experiments, through a behavioral prediction task and computational modeling, we report evidence of humans dissociating volatility and stochasticity solely based on their observations. We observed contrasting effects of volatility and stochasticity on learning rates, consistent with statistical principles. These results are consistent with a computational model that estimates volatility and stochasticity by balancing their dueling effects.

Repeating patterns: Predictive processing suggests an aesthetic learning role of the basal ganglia in repetitive stereotyped behaviors

Recurrent, unvarying, and seemingly purposeless patterns of action and cognition are part of normal development, but also feature prominently in several neuropsychiatric conditions. Repetitive stereotyped behaviors (RSBs) can be viewed as exaggerated forms of learned habits and frequently correlate with alterations in motor, limbic, and associative basal ganglia circuits. However, it is still unclear how altered basal ganglia feedback signals actually relate to the phenomenological variability of RSBs. Why do behaviorally overlapping phenomena sometimes require different treatment approaches-for example, sensory shielding strategies versus exposure therapy for autism and obsessive-compulsive disorder, respectively? Certain clues may be found in recent models of basal ganglia function that extend well beyond action selection and motivational control, and have implications for sensorimotor integration, prediction, learning under uncertainty, as well as aesthetic learning. In this paper, we systematically compare three exemplary conditions with basal ganglia involvement, obsessive-compulsive disorder, Parkinson's disease, and autism spectrum conditions, to gain a new understanding of RSBs. We integrate clinical observations and neuroanatomical and neurophysiological alterations with accounts employing the predictive processing framework. Based on this review, we suggest that basal ganglia feedback plays a central role in preconditioning cortical networks to anticipate self-generated, movement-related perception. In this way, basal ganglia feedback appears ideally situated to adjust the salience of sensory signals through precision weighting of (external) new sensory information, relative to the precision of (internal) predictions based on prior generated models. Accordingly, behavioral policies may preferentially rely on new data versus existing knowledge, in a spectrum spanning between novelty and stability. RSBs may then represent compensatory or reactive responses, respectively, at the opposite ends of this spectrum. This view places an important role of aesthetic learning on basal ganglia feedback, may account for observed changes in creativity and aesthetic experience in basal ganglia disorders, is empirically testable, and may inform creative art therapies in conditions characterized by stereotyped behaviors.

Emotion dynamics as hierarchical Bayesian inference in time

What fundamental property of our environment would be most valuable and optimal in characterizing the emotional dynamics we experience in daily life? Empirical work has shown that an accurate estimation of uncertainty is necessary for our optimal perception, learning, and decision-making. However, the role of this uncertainty in governing our affective dynamics remains unexplored. Using Bayesian encoding, decoding and computational modeling, on a large-scale neuroimaging and behavioral data on a passive movie-watching task, we showed that emotions naturally arise due to ongoing uncertainty estimations about future outcomes in a hierarchical neural architecture. Several prefrontal subregions hierarchically encoded a lower-dimensional signal that highly correlated with the evolving uncertainty. Crucially, the lateral orbitofrontal cortex (lOFC) tracked the temporal fluctuations of this uncertainty and was predictive of the participants' predisposition to anxiety. Furthermore, we observed a distinct functional double-dissociation within OFC with increased connectivity between medial OFC and DMN, while with that of lOFC and FPN in response to the evolving affect. Finally, we uncovered a temporally predictive code updating an individual's beliefs spontaneously with fluctuating outcome uncertainty in the lOFC. A biologically relevant and computationally crucial parameter in the theories of brain function, we propose uncertainty to be central to the definition of complex emotions.

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

Delusions are rigid beliefs held with high certainty despite contradictory evidence. Notwithstanding decades of research, we still have a limited understanding of the computational and neurobiological alterations giving rise to delusions. In this review, we highlight a selection of recent work in computational psychiatry aimed at developing quantitative models of inference and its alterations, with the goal of providing an explanatory account for the form of delusional beliefs in psychosis. First, we assess and evaluate the experimental paradigms most often used to study inferential alterations in delusions. Based on our review of the literature and theoretical considerations, we contend that classic draws-to-decision paradigms are not well-suited to isolate inferential processes, further arguing that the commonly cited 'jumping-to-conclusion' bias may reflect neither delusion-specific nor inferential alterations. Second, we discuss several enhancements to standard paradigms that show promise in more effectively isolating inferential processes and delusion-related alterations therein. We further draw on our recent work to build an argument for a specific failure mode for delusions consisting of prior overweighting in high-level causal inferences about partially observable hidden states. Finally, we assess plausible neurobiological implementations for this candidate failure mode of delusional beliefs and outline promising future directions in this area.

Predictors of short-term impulsive and compulsive behaviour after subthalamic stimulation in Parkinson disease

BACKGROUND

The effects of subthalamic stimulation (subthalamic nucleus-deep brain stimulation, STN-DBS) on impulsive and compulsive behaviours (ICB) in Parkinson's disease (PD) are understudied.

OBJECTIVE

To investigate clinical predictors of STN-DBS effects on ICB.

METHODS

In this prospective, open-label, multicentre study in patients with PD undergoing bilateral STN-DBS, we assessed patients preoperatively and at 6-month follow-up postoperatively. Clinical scales included the Questionnaire for Impulsive-Compulsive Disorders in PD-Rating Scale (QUIP-RS), PD Questionnaire-8, Non-Motor Symptom Scale (NMSS), Unified PD Rating Scale in addition to levodopa-equivalent daily dose total (LEDD-total) and dopamine agonists (LEDD-DA). Changes at follow-up were analysed with Wilcoxon signed-rank test and corrected for multiple comparisons (Bonferroni method). We explored predictors of QUIP-RS changes using correlations and linear regressions. Finally, we dichotomised patients into 'QUIP-RS improvement or worsening' and analysed between-group differences.

RESULTS

We included 55 patients aged 61.7 years±8.4 with 9.8 years±4.6 PD duration. QUIP-RS cut-offs and psychiatric assessments identified patients with preoperative ICB. In patients with ICB, QUIP-RS improved significantly. However, we observed considerable interindividual variability of clinically relevant QUIP-RS outcomes as 27.3% experienced worsening and 29.1% an improvement. In post hoc analyses, higher baseline QUIP-RS and lower baseline LEDD-DA were associated with greater QUIP-RS improvements. Additionally, the 'QUIP-RS worsening' group had more severe baseline impairment in the NMSS attention/memory domain.

CONCLUSIONS

Our results show favourable ICB outcomes in patients with higher preoperative ICB severity and lower preoperative DA doses, and worse outcomes in patients with more severe baseline attention/memory deficits. These findings emphasise the need for comprehensive non-motor and motor symptoms assessments in patients undergoing STN-DBS.

TRIAL REGISTRATION NUMBER

DRKS00006735.

A model for learning based on the joint estimation of stochasticity and volatility

Previous research has stressed the importance of uncertainty for controlling the speed of learning, and how such control depends on the learner inferring the noise properties of the environment, especially volatility: the speed of change. However, learning rates are jointly determined by the comparison between volatility and a second factor, moment-to-moment stochasticity. Yet much previous research has focused on simplified cases corresponding to estimation of either factor alone. Here, we introduce a learning model, in which both factors are learned simultaneously from experience, and use the model to simulate human and animal data across many seemingly disparate neuroscientific and behavioral phenomena. By considering the full problem of joint estimation, we highlight a set of previously unappreciated issues, arising from the mutual interdependence of inference about volatility and stochasticity. This interdependence complicates and enriches the interpretation of previous results, such as pathological learning in individuals with anxiety and following amygdala damage.

The structural connectivity of subthalamic deep brain stimulation correlates with impulsivity in Parkinson's disease

Subthalamic deep brain stimulation (STN-DBS) for Parkinson's disease treats motor symptoms and improves quality of life, but can be complicated by adverse neuropsychiatric side-effects, including impulsivity. Several clinically important questions remain unclear: can 'at-risk' patients be identified prior to DBS; do neuropsychiatric symptoms relate to the distribution of the stimulation field; and which brain networks are responsible for the evolution of these symptoms? Using a comprehensive neuropsychiatric battery and a virtual casino to assess impulsive behaviour in a naturalistic fashion, 55 patients with Parkinson's disease (19 females, mean age 62, mean Hoehn and Yahr stage 2.6) were assessed prior to STN-DBS and 3 months postoperatively. Reward evaluation and response inhibition networks were reconstructed with probabilistic tractography using the participant-specific subthalamic volume of activated tissue as a seed. We found that greater connectivity of the stimulation site with these frontostriatal networks was related to greater postoperative impulsiveness and disinhibition as assessed by the neuropsychiatric instruments. Larger bet sizes in the virtual casino postoperatively were associated with greater connectivity of the stimulation site with right and left orbitofrontal cortex, right ventromedial prefrontal cortex and left ventral striatum. For all assessments, the baseline connectivity of reward evaluation and response inhibition networks prior to STN-DBS was not associated with postoperative impulsivity; rather, these relationships were only observed when the stimulation field was incorporated. This suggests that the site and distribution of stimulation is a more important determinant of postoperative neuropsychiatric outcomes than preoperative brain structure and that stimulation acts to mediate impulsivity through differential recruitment of frontostriatal networks. Notably, a distinction could be made amongst participants with clinically-significant, harmful changes in mood and behaviour attributable to DBS, based upon an analysis of connectivity and its relationship with gambling behaviour. Additional analyses suggested that this distinction may be mediated by the differential involvement of fibres connecting ventromedial subthalamic nucleus and orbitofrontal cortex. These findings identify a mechanistic substrate of neuropsychiatric impairment after STN-DBS and suggest that tractography could be used to predict the incidence of adverse neuropsychiatric effects. Clinically, these results highlight the importance of accurate electrode placement and careful stimulation titration in the prevention of neuropsychiatric side-effects after STN-DBS.

Machine Learning's Application in Deep Brain Stimulation for Parkinson's Disease: A Review

Deep brain stimulation (DBS) is a surgical treatment for advanced Parkinson's disease (PD) that has undergone technological evolution that parallels an expansion in clinical phenotyping, neurophysiology, and neuroimaging of the disease state. Machine learning (ML) has been successfully used in a wide range of healthcare problems, including DBS. As computational power increases and more data become available, the application of ML in DBS is expected to grow. We review the literature of ML in DBS and discuss future opportunities for such applications. Specifically, we perform a comprehensive review of the literature from PubMed, the Institute for Scientific Information's Web of Science, Cochrane Database of Systematic Reviews, and Institute of Electrical and Electronics Engineers' (IEEE) Xplore Digital Library for ML applications in DBS. These studies are broadly placed in the following categories: (1) DBS candidate selection; (2) programming optimization; (3) surgical targeting; and (4) insights into DBS mechanisms. For each category, we provide and contextualize the current body of research and discuss potential future directions for the application of ML in DBS.

Deep brain stimulation and recordings: Insights into the contributions of subthalamic nucleus in cognition

Recent progress in targeted interrogation of basal ganglia structures and networks with deep brain stimulation in humans has provided insights into the complex functions the subthalamic nucleus (STN). Beyond the traditional role of the STN in modulating motor function, recognition of its role in cognition was initially fueled by side effects seen with STN DBS and later revealed with behavioral and electrophysiological studies. Anatomical, clinical, and electrophysiological data converge on the view that the STN is a pivotal node linking cognitive and motor processes. The goal of this review is to synthesize the literature to date that used DBS to examine the contributions of the STN to motor and non-motor cognitive functions and control. Multiple modalities of research have provided us with an enhanced understanding of the STN and reveal that it is critically involved in motor and non-motor inhibition, decision-making, motivation and emotion. Understanding the role of the STN in cognition can enhance the therapeutic efficacy and selectivity not only for existing applications of DBS, but also in the development of therapeutic strategies to stimulate aberrant circuits to treat non-motor symptoms of Parkinson's disease and other disorders.

The structural connectivity of discrete networks underlies impulsivity and gambling in Parkinson’s disease

See O’Callaghan (doi:10.1093/brain/awz349) for a scientific commentary on this article.

Mosley et al. examine impulsivity and naturalistic gambling behaviours in patients with Parkinson’s disease. They link within-patient differences to the structural connectivity of networks subserving reward evaluation and response inhibition, and reveal pivotal roles for the ventral striatum and subthalamic nucleus within these networks.