Multi-modal MRI data fusion to assess neurobiological effects of marijuana use

描述（由申请人提供）：2012年监测未来调查发现，6.5%的高中生每天吸食大麻，高于五年前的5.1%。与此同时，只有44.1%的年龄较大的青少年认为经常使用大麻是有害的（自1979年以来的最低水平）。由于感知伤害是使用的关键指标，并且大麻合法化运动在美国取得了成功，这些趋势将继续下去，并可能导致新的健康危机。虽然关于MJ危害的辩论激烈，但神经影像学研究尚未充分发挥其潜力，为关于这一关键社会问题的全国对话提供信息。例如，磁共振成像（MRI）研究表明，与MJ使用相关的大脑结构和功能存在一些差异，然而，研究结果是模棱两可的。其中一个可能的原因是，MJ使用的影响与个人的特定药物使用模式交织在一起，以至于很难使用传统的统计方法来确定MJ使用者和对照组之间的差异（因为MJ使用者的差异很大）。为了解决这一差距，该项目结合了来自五个不同的NIDA资助的慢性MJ使用研究的MRI数据，以从更大的MJ使用者样本中获益。一种新的统计方法，只适用于大数据集，称为数据融合，然后将应用于组合数据集。数据融合利用了MJ使用可能产生的广泛影响，而不是被它混淆。这种方法的另一个优点是，它整合了多个MRI测量结果的信息，将灰色和白色物质结构与大脑网络的回路级行为联系在一起。对现有数据的分析将侧重于调查MJ使用对MJ使用者执行功能和认知控制的影响，具体如下：1）评估在认知任务期间协调在一起的大规模大脑网络之间的相互作用，2）应用数据融合方法将认知网络功能与大脑结构联系起来，进而解开长期使用MJ的影响，以及3）最重要的是，在另一个大型数据集中测试这些发现的预测价值，该数据集将在本研究的第三年提供。在新样本中验证结构-功能模式与MJ使用之间的关系是成瘾神经成像的一个重大进步，这将导致新的生物标志物以检测脆弱个体为目标，并提出新的诊断，预防和治疗策略。与NIDA的治疗和预防战略目标相一致，拟议的研究将：1）促进我们对长期使用大麻对大脑结构和网络电路的神经生物学影响的理解，2）使用现有数据的新统计分析结合不同数据集的验证来确定生物标志物，以表征脆弱的个体，并提出新的诊断，预防和治疗策略。

DESCRIPTION (provided by applicant): The 2012 Monitoring the Future survey found that 6.5% of high school seniors smoke marijuana daily, up from 5.1% from five years ago. At the same time, only 44.1% of older teens see regular marijuana use as harmful (the lowest since 1979). Because perceived harm is a key indicator of use and movements to legalize marijuana (MJ) are succeeding in the US, these trends will continue and may lead to a new health crisis. While the debate as to the harms of MJ rages, neuroimaging studies have not realized their full potential to inform the national dialogue on this key social issue. For example, magnetic resonance imaging (MRI) studies suggest some differences in brain structure and function associated with MJ use, however, the findings are equivocal. One possible reason for this is that the effects of MJ use are so intertwined with specific drug use patterns in the individual that it becomes difficult to identify differences between MJ users and controls using conventional statistical methods (because there is such large variability in MJ users). To address this gap, this project combines MRI data from five different NIDA-funded studies of chronic MJ use to derive benefit from having a much larger sample of MJ users. A new statistical method suitable only with large datasets, called data fusion, will then be applied to the combined dataset. Data fusion capitalizes on the wide range of effects that MJ use may engender, rather than being confounded by it. A further strength of this method is that it integrates information across multiple MRI measurements to link together, for example, gray and white matter structure with circuit-level behavior of brain networks. Analyses of extant data will focus on investigating the effects of MJ use on executive function and cognitive control in MJ users by: 1) assessing interactions between large-scale brain networks that coordinate together during cognitive tasks, 2) applying the data fusion approach to link cognitive network function with brain structure and, in turn, disentangle the impact of chronic MJ use, and 3) most importantly, testing the predictive value of these findings in another large dataset that will be available in the third year of this study. Validating relationships between structure-function patterns and MJ use in a new sample is a significant advancement to addiction neuroimaging that will lead to novel biomarkers to target the detection of vulnerable individuals and to suggest new diagnostic, prevention and treatment strategies. Consistent with NIDA's strategic goals of treatment and prevention, the proposed research will: 1) Advance our understanding of the neurobiological effects of chronic marijuana use on brain structure and network circuitry, 2) Use novel statistical analyses of extant data combined with validation in a different dataset to identify biomarkers to characterize vulnerable individuals and suggest new diagnostic, prevention, and treatment strategies.