Bridging Scales to Understand Endogenous Neuromodulation and its Regulation

点击翻译按钮获取中文摘要

Neuromodulatory nuclei detect and transform brain network activity into simpler signals, then send neurotransmitters back out to large-scale brain networks to change their function. Such nuclei are centrally implicated in mental disorders and adaptive resilience, and their regulation remains an untapped resource for interventions. The purpose of this grant is to understand how neuromodulatory nuclei detect and in turn influence distributed patterns of brain activity to impact behavior. To understand their regulation and effects on brain function, the investigative team has developed novel neuroimaging, behavioral, and analytic methods. These methods include: training participants to endogenously self-regulate dopaminergic midbrain, isolating distinct streams of information in the midbrain over multiple timescales, distinguishing behavioral contexts and network effects associated with univariate activation in neuromodulatory nuclei, and finally relating midbrain activation to memory-conducive states in medial temporal lobe memory systems. Our team has recently developed whole-brain analyses of real-time fMRI during midbrain neurofeedback and machine-learning tools for characterizing nonlinear latent dynamics from high-dimensional data. Now, with these tools, we can relate midbrain activation to whole brain states. We hypothesize 1) that distinct distributed spatiotemporal patterns precede and follow midbrain univariate activation, specify it uniquely among neuromodulatory nuclei, and distinguish sustained from transient midbrain responses; 2) that the evolution of these patterns over the training session will predict learning to upregulate midbrain, and 3) that endogenous midbrain regulation will predict brain and behavioral effects we and others have previously shown to be associated with midbrain activation and dopamine function. If the aims of this project are achieved, we will have introduced a multi-level model of the neural states that support midbrain activation, a complement of methods for regulating midbrain noninvasively, and an improved understanding of its impact on learning and motivated behavior. Reliable cognitive strategies for dynamically and selectively fine-tuning neural networks to suit behavioral contexts will lay the foundation for a wide array of interventions across educational and clinical applications.