Inter-regional interactions underlying the decision to act, at spiking resolution

The adaptive value of cognition manifests in decisions not only about how to act but also about when to act. Though much neuroscientific investigation of decision-making has focused on sensory-guided perceptual decisions, somewhat distinct neural processes appear to underlie internally generated (self-driven) decisions of when to act. After early electroencephalographic recordings in humans identified a ‘readiness potential’ that precedes the initiation of self-driven movements, numerous cortical and subcortical regions across different mammalian models were functionally implicated in the self-driven decision to act. However, the historical challenge of assessing interactions between these regions at the temporal resolution of spiking has prevented functional ideas from being adequately tested, keeping mechanistic understanding out of reach. Our collaboration aims to unify the disparate literature on regions functionally implicated in self-driven action decisions and to clarify relevant neural activity dynamics across these regions. We will use a range of emerging experimental and computational methods to develop network models that explain the decision to act through interactions between neuronal populations on the timescale of synaptic communication. We will use a climbing paradigm for head-fixed mice that we have recently developed, leveraging the self-driven decision to initiate climbing that naturally emerges in this paradigm. We will employ a comparative task design that pairs self-driven, cue-based, and reward-contingent action decisions, allowing us to distinguish neural activity underlying action decision, initiation and movement planning. Simultaneous multi-array, multi-region recording across implicated regions will allow us to quantify interactions between neuronal populations with cellular and spike resolution. Characterization of neural activity dynamics will help unify models, clarify the flow of decision-related activity, and expose relevant network architectures. We will pair contemporary approaches for online detection of population activity events and optogenetics to causally test and iteratively refine resulting network models. We expect our work to radically improve mechanistic understanding of self-driven action decisions, a fundamental and long-studied aspect of cognition.