Brain Network Mechanisms of Instructed Learning

点击翻译按钮获取中文摘要

Project Summary/Abstract: The tools of network science have enabled substantial progress in understanding the intrinsic organization of the human brain. Yet, the predominant focus on resting-state functional connectivity (FC) has become a critical barrier to progress in cognitive neuroscience, given that rest FC does not account for task-specific network changes likely essential for adaptive cognition. We offer a complementary approach – cognitive network neuroscience – which applies dynamic network analysis tools and theories to task manipulations of FC to offer insights into human cognitive function. The goal of this proposal is to utilize this network-based approach with human neuroimaging to understand how instructed learning is implemented in the human brain, from initial learning to automaticity after extensive practice. Most neuroscientific research has focused on non-instructed (e.g., exploratory or feedback-based) learning. Yet, instructed learning is highly relevant to mental health for several reasons. First, empirically supported psychotherapies (e.g., cognitive behavioral therapy) utilize the human ability for rapid instructed task learning (RITL; “rittle”) to convert instructions into cognitive strategies that improve outcomes across nearly every major mental disease. Second, RITL is impaired in a variety of mental diseases, given that RITL relies on flexible cognitive control – a general capacity supporting adaptive, goal-directed behavior important in daily life. Thus, in addition to adding difficulties to everyday life (e.g., learning new skills at work), the disruption of RITL abilities likely limits the effectiveness of psychotherapy in improving mental health. Finally, instructed learning provides an especially powerful means of experimental control over behavior change, which underlies mental health improvements even outside the context of psychotherapy. Advancing understanding of the neural basis of RITL and its transition to practiced automatized behaviors parallels the transition from instructions in the clinic to ingrained habits that can foster successful mental health change. In prior work, we built a large-scale brain network theory for how instructed learning occurs by drawing on the concept of “flexible hubs” – brain regions that coordinate goal-directed cognition (flexible control) by dynamically updating connectivity throughout the brain. The flexible hub theory strongly links the methods and theories of network science to the cognitive neuroscience of learning, and as such has the power to offer insights into the large-scale network processes underlying instructed learning. We propose to use large-scale brain network theory to understand the domain generality of flexible hubs during instructed learning (Aim 1), to determine the role of flexible hubs in the transition from novel instructed task training to practiced performance (Aims 2.1 & 2.2), and to develop RITL cognitive training that maximizes the utility of flexible hubs for performance of novel tasks (Aim 2.3). Our network-based approach to understanding instructed learning along with RITL cognitive training may lead to improved outcomes for a variety of mental disorders, given the central role instructed learning plays in empirically supported psychotherapies.