Computational models of behavioral addictions: State of the art and future directions

Kato, Ayaka; Shimomura, Kanji; Ognibene, Dimitri; Parvaz, Muhammad A.; Berner, Laura A.; Morita, Kenji; Fiore, Vincenzo

doi:10.1016/j.addbeh.2022.107595

Cited by 12 publications

(7 citation statements)

References 111 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The network dynamic measures derived here are consistent with frameworks using concepts from nonlinear dynamical systems to investigate affective experiences (47, 48), psychopathology (49, 50), and substance use (51, 52). Following these studies, we may also consider the PND and NND states as attractors (i.e., stable states the brain can achieve).…”

Section: Discussionsupporting

confidence: 71%

Network state dynamics underpin craving in a transdiagnostic population

Ye,

Garrison,

Lacadie

et al. 2023

Preprint

View full text Add to dashboard Cite

Emerging fMRI brain dynamic methods present a unique opportunity to capture how brain region interactions across time give rise to evolving affective and motivational states. As the unfolding experience and regulation of affective states affect psychopathology and well-being, it is important to elucidate their underlying time-varying brain responses. Here, we developed a novel framework to identify network states specific to an affective state of interest and examine how their instantaneous engagement contributed to its experience. This framework investigated network state dynamics underlying craving, a clinically meaningful and changeable state. In a transdiagnostic sample of healthy controls and individuals diagnosed with or at risk for craving-related disorders (N=252), we utilized connectome-based predictive modeling (CPM) to identify craving-predictive edges. An edge-centric timeseries approach was leveraged to quantify the instantaneous engagement of the craving-positive and craving-negative networks during independent scan runs. Individuals with higher craving persisted longer in a craving-positive network state while dwelling less in a craving-negative network state. We replicated the latter results externally in an independent group of healthy controls and individuals with alcohol use disorder exposed to different stimuli during the scan (N=173). The associations between craving and network state dynamics can still be consistently observed even when craving-predictive edges were instead identified in the replication dataset. These robust findings suggest that variations in craving-specific network state recruitment underpin individual differences in craving. Our framework additionally presents a new avenue to explore how the moment-to-moment engagement of behaviorally meaningful network states supports our changing affective experiences.

show abstract

Section: Discussionsupporting

confidence: 71%

Network state dynamics underpin craving in a transdiagnostic population

Ye,

Garrison,

Lacadie

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…These studies have demonstrated reduced cognitive control and high impulsivity levels in these individuals [31][32][33][34] . Based on this literature, one might expect that smokers would exhibit reduced planning horizon as suggested by previous computational work 6,35 . Here, formal model comparison showed that smokers engaged a similar 2-step forward thinking model as non-smoking controls, yet under-estimated the in uence of their actions on future states (lower value) compared to non-smokers.…”

Section: Discussionmentioning

confidence: 80%

Aberrant neural computation of social controllability in nicotine-dependent humans

Gu,

McLaughlin,

et al. 2024

Preprint

View full text Add to dashboard Cite

Social controllability, defined as the ability to exert influence when interacting with others, is crucial for optimal decision-making. Inability to do so might contribute to maladaptive behaviors such as drug use, which often takes place in social settings. Here, we examined nicotine-dependent humans using fMRI, as they made choices that could influence the proposals from simulated partners. Computational modeling revealed that smokers under-estimated the influence of their actions and self-reported a reduced sense of control, compared to non-smokers. These findings were replicated in a large independent sample of participants recruited online. Neurally, smokers showed reduced tracking of forward projected choice values in the ventromedial prefrontal cortex, and impaired computation of social prediction errors in the midbrain. These results demonstrate that smokers were less accurate in estimating their personal influence when the social environment calls for control, providing a neurocomputational account for the social cognitive deficits in this population.

show abstract

“…The development of neurocomputational models has enabled us to understand the cognitive mechanisms involved in decision-making. Current findings have revealed that models relying on reinforcement learning (RL) algorithms, and Bayesian inference, which focus on vulnerabilities related to model-free and model-based control, can explain maladaptive choices despite adverse consequences in behavioral tasks such as the Iowa Gambling Task (IGT) (Lin et al, 2019;Kato et al, 2023). In the current study, we aimed to explore reward-processing biases using a simpler agent model without introducing additional parameters.…”

Section: Discussionmentioning

confidence: 99%

Computational modeling of decision-making in substance abusers: testing Bechara’s hypotheses

Avila Chauvet,

Mejía Cruz

2024

Front. Psychol.

View full text Add to dashboard Cite

One of the cognitive abilities most affected by substance abuse is decision-making. Behavioral tasks such as the Iowa Gambling Task (IGT) provide a means to measure the learning process involved in decision-making. To comprehend this process, three hypotheses have emerged: (1) participants prioritize gains over losses, (2) they exhibit insensitivity to losses, and (3) the capacity of operational storage or working memory comes into play. A dynamic model was developed to examine these hypotheses, simulating sensitivity to gains and losses. The Linear Operator model served as the learning rule, wherein net gains depend on the ratio of gains to losses, weighted by the sensitivity to both. The study further proposes a comparison between the performance of simulated agents and that of substance abusers (n = 20) and control adults (n = 20). The findings indicate that as the memory factor increases, along with high sensitivity to losses and low sensitivity to gains, agents prefer advantageous alternatives, particularly those with a lower frequency of punishments. Conversely, when sensitivity to gains increases and the memory factor decreases, agents prefer disadvantageous alternatives, especially those that result in larger losses. Human participants confirmed the agents’ performance, particularly when contrasting optimal and sub-optimal outcomes. In conclusion, we emphasize the importance of evaluating the parameters of the linear operator model across diverse clinical and community samples.

show abstract

Computational models of behavioral addictions: State of the art and future directions

Cited by 12 publications

References 111 publications

Network state dynamics underpin craving in a transdiagnostic population

Network state dynamics underpin craving in a transdiagnostic population

Aberrant neural computation of social controllability in nicotine-dependent humans

Computational modeling of decision-making in substance abusers: testing Bechara’s hypotheses

Contact Info

Product

Resources

About