Interdisciplinary studies of decision-making and motor planning

大多数决策理论认为，大脑首先根据主观价值、成本、风险等因素构建抽象的选择表征，然后用这些来选择目标，最后将目标转化为行动命令。然而，越来越多的神经数据使这种经典的系列框架越来越受到质疑。我相信，部分原因是大脑是漫长进化历史的产物，在这一过程中，大多数动物都面临着与抽象认知截然不同的挑战。例如，动物生活中的大多数决定都是在实时的、动态变化的环境中做出的，这就没有时间进行仔细的考虑，然后制定行动计划。为了在这种动态的情况下有效的行为，大脑需要一个并行的架构，在这个架构中，多个潜在的动作可以同时被指定，其中一个是根据不断变化的感官信息选择的，而其他的选择在情况突然变化的情况下继续被考虑。为了处理这种情况，我提出了“功能竞争假说”，该假说认为，行为的功能组织是一种平行结构，在这种结构中，感觉信息被处理以指定在感觉运动回路中相互竞争的几个潜在动作，而其他系统收集信息以使竞争偏向于选择获胜的动作。我的研究计划侧重于测试和进一步发展这一广泛的理论框架。我通过一种跨学科的方法来解决这些目标，该方法将计算建模与一系列实验技术相结合，包括在清醒的、行为的猴子中进行多电极记录，以及人类行为、成像和经颅磁刺激研究。我的NSERC发现基金支持计算项目以及人类受试者的实验研究。目前的项目分为三个主题。主题#1阐述了大脑如何在动态变化的环境中做出决策，并将涉及基于cihr资助的猴子神经记录实验和nserc资助的人类脑磁图（MEG）的现有数据开发一个计算模型。主题2将这些想法推广到更高层次的决策中，但仍然在自然环境中框架。特别是，我们将在人类身上开展一系列新的脑脑电信号研究，以调查在参与当前可用的机会与在其他地方寻找新机会之间的“觅食”选择。主题#3涉及一个非常新颖的研究方向：当我们已经采取行动时，决策是如何做出的。这建立在我们在之前的发现基金资助周期中开创的行为范式的基础上，我们将通过在各种手动跟踪任务中使用经颅磁刺激来研究人类的神经机制。

Most theories of decision-making suggest that the brain first builds abstract representations of choices in terms of subjective values, costs, risks, etc., then uses these to select a goal, and finally transforms that goal into a movement command. However, a growing body of neural data increasingly casts such classical serial frameworks into doubt. Part of the reason, I believe, is that the brain is the product of a long history of evolution, during most of which animals were faced with challenges very different from those of abstract cognition. For example, most decisions in animal life are made in a real-time, dynamically changing environment, which does not allow time for careful deliberation followed by action planning. For effective behavior in such dynamic situations, the brain requires a parallel architecture in which multiple potential actions can be specified simultaneously, one selected on the basis of continuously changing sensory information, and all the while other options continue to remain under consideration in case circumstances suddenly change. To deal with such situations, I've proposed the "affordance competition hypothesis", which suggests that the functional organization of behavior is a parallel architecture whereby sensory information is processed to specify several potential actions that compete against each other in sensorimotor circuits while other systems collect information for biasing that competition to select a winning action. My research program focuses on testing and further developing this broad theoretical framework. I address these goals through an interdisciplinary approach that integrates computational modeling with a range of experimental techniques, including multi-electrode recording in awake, behaving monkeys, as well as human behavioral, imaging, and transcranial magnetic stimulation studies. My NSERC Discovery Grant supports the computational projects as well as experimental studies with human subjects. Current projects fall into three general themes. Theme #1 addresses how the brain makes decisions in a dynamically changing environment, and will involve developing a computational model on the basis of existing data from CIHR-funded neural recording experiments in monkeys as well as NSERC-funded magnetoencephalography (MEG) in humans. Theme #2 generalizes these ideas toward higher-level decisions, but still framed within a natural context. In particular, we will launch a new series of MEG studies in humans to investigate "foraging" choices between engaging currently available opportunities versus searching for new opportunities elsewhere. Theme #3 addresses a very novel line of research: how decisions are made while we are already acting. This builds on behavioral paradigms we pioneered in the previous funding cycle of my Discovery Grant, which we will extend toward investigating neural mechanisms through the use of transcranial magnetic stimulation in humans during a variety of manual tracking tasks.