Neurocomputational Mechanisms for Addiction Heterogeneity, Impulsivity and Perseverance

项目摘要 对物质使用障碍(Suds)的研究已经确定了一种深刻的学科间变异性，其中广泛的 多种多方面的、可分离的行为表型与成瘾发展和 症状学。即使是明显不相容的行为表达，如冲动和缺乏弹性 或毅力，已被发现在肥皂水中共同出现，并预示着类似的成瘾弱点。 然而，到目前为止，很少有神经或计算机制被描述来解释这种情况 看似矛盾的发现和个体差异，从而阻碍了 个体化诊疗。这个项目的总体目标是验证一个新的模型 使用尼古丁成瘾使用障碍(NUD)作为测试案例。我们建议对以前的理论进行扩展，以 提供更全面的成瘾神经计算框架，包括表型 冲动和毅力的变异性和共生性，以有效率为特征 皮质纹状体环路的连通性。这个项目的科学前提是几十年来 人类和非人类动物的研究已经证明了腹侧在成瘾中所起的作用 和背侧皮质纹状体系统，分别负责目标导向和习惯性行为。这个 对这两个回路的神经动力学的模拟使我们的模型能够描述两个回路上的成瘾 独立的维度。在第一个维度上，尼古丁等成瘾药物会导致回路增加 腹侧和背侧皮质-纹状体系统的增益和状态转换稳定性，放大初步 证据(冲动)和做出选择由于反馈效应而变得缺乏弹性 (毅力)。在第二个维度上，我们的模型不一定会受到药物暴露的影响 一致认为，这两个皮质纹状体回路与增益相关的过度稳定是由 两条赛道中的任何一条都存在对另一条赛道的“支配”。在目标1中，我们将验证该模型 预测高电路增益预示着更强的行为冲动和毅力。在《目标2》中我们将 验证两个皮质-纹状体回路之间的平衡预测药物使用的模型预测 严肃性。电路增益和电路平衡将在NUD患者(n=32)和健康对照组中进行测试 (n=32)，负责决策任务。电路增益将用有效连通性来衡量 皮质和纹状体区域之间，在每个回路内，并使用动态因果关系进行估计 建模(DCM)。将使用DCM估计腹背有效连接的电路平衡， 在坡度上建立统治地位。该概念验证项目可以提供新的计算能力 药物成瘾的框架，以及描述临床异质性的量化模型，最终 告知个体化治疗。

Project Summary Studies in substance use disorders (SUDs) have identified a profound inter-subject variability, where a wide variety of multifaceted, dissociable behavioral phenotypes are correlated with addiction development and symptomatology. Even apparently incompatible behavioral expressions, such as impulsivity and inelasticity or perseverance, have been found to co-occur in SUDs and to signal similar addiction vulnerabilities. However, few neural or computational mechanisms have been described so far to account for such seemingly contradicting findings and individual differences, thus hindering the development of individualized diagnosis and treatment. The overarching goal of this project is to validate a new model of addiction using Nicotine Use Disorder (NUD) as test case. We propose to expand on previous theories to provide a more comprehensive neuro-computational framework of addiction that includes phenotypic variability and co-occurrence of impulsivity and perseverance, characterized in terms of effective connectivity in cortico-striatal circuits. The scientific premise for this project is grounded in decades of human and non-human animal work which have demonstrated the roles played in addiction by the ventral and dorsal corticostriatal systems, respectively responsible for goal oriented and habitual behavior. The simulation of the neural dynamics in these two circuits has allowed our model to describe addiction on two independent dimensions. On a first dimension, addictive drugs such as nicotine result in increased circuit gain and state transition stability in both ventral and dorsal cortico-striatal systems, amplifying preliminary evidence (impulsivity) and making choice selections become inelastic due to a feedback effect (perseverance). On a second dimension, which is not necessarily affected by drug exposure, our models converge in suggesting that this gain-related over-stability of both cortico-striatal circuits is aggravated by the presence of a “dominance” of either of the two circuit over the other. In aim 1, we will validate the model prediction that high circuit gain predicts greater behavioral impulsivity and perseverance. In aim 2 we will validate the model prediction that the balance between the two cortico-striatal circuits predicts drug use severity. Circuit gain and circuit balance will be tested in NUD individuals (n=32) and healthy controls (n=32), tasked with decision-making tasks. Circuit gain will be measured in terms of effective connectivity between cortical and striatal areas, within each circuit, and estimated with the use of Dynamic Causal Modelling (DCM). Circuit balance will be estimated using DCM for the ventro-dorsal effective connectivity, to establish dominance on a gradient. This proof-of-concept project can provide a new computational framework for drug addiction, and a quantitative model to characterize clinical heterogeneity, eventually informing individualized treatments.