Decision-making studies often focus on brain mechanisms for selecting between choices. However, another important, and often neglected, aspect of decision-making concerns whether, at any given point in time, it is worth taking any action at all. By performing a series of experiments in humans and non-human primates we investigated the neural mechanisms that underlie free voluntary action or that translate a desire into an action. First, we showed how macaques made decisions about when it is worth acting as a result of activity patterns that emerge in and between the anterior cingulate cortex (ACC) and the basal forebrain (BF). Activity in these structures mediated the influence of environmental features, such as its potential for reward, on whether an action will be emitted by the macaque. Second, we used a novel brain stimulation technique, transcranial ultrasound stimulation, to show the causal importance of ACC and BF in decisions about when to act. Third, by using ultra-high field functional magnetic resonance imaging (7T fMRI) we demonstrated the importance of the basal forebrain-cingulate circuit in humans and identified a multi-layered circuit in the human brain originating far beyond the medial frontal areas typically linked to human voluntary action initiation starting in the basal forebrain, pedunculopontine nucleus, and habenula, converging in the substantia nigra, and only then projecting to striatum and cortex.

Although decision-making has traditionally been assumed to be distinct from movement execution, for many years a growing body of data has reported neural correlates of decision variables in many regions associated with motor control, casting doubt on that assumption. In this presentation, I’ll show evidence that decision-making and action execution are two modes of a unified dynamical system implemented in large distributed populations of neurons. We recorded neurons in the dorsal premotor cortex, the primary motor cortex, the dorsolateral prefrontal cortex and in the external and internal segments of the globus pallidus of two monkeys trained to perform a probabilistic reach decision task in which sensory evidence evolves within each trial. With “state-space” analyses, I’ll show how these areas share labor during the decision process and how activity patterns are compatible with a dynamical attractor model in which cortical activity reflects a biased competition between actions, gradually amplified by an urgency signal from the basal ganglia.

TBC

TBC

TBC

TBC

To interact meaningfully with its environment, an agent must integrate external information with its own internal states. In a set of two studies, we identified a neural correlate that tracks how categorical choice unfolds as sequential evidence is sampled in various urgency conditions. In our task participants had to monitor long sequences of discrete visual stimuli and decide which out of two sets of stimuli was most frequent. In a first experiment, we found that the classic P3 event-related potential evoked by sequential items tracked decision formation during evidence accumulation and predicted participants’ categorical choices on a single trial level, both when evidence was strong and when it was ambiguous. Participants’ reaction times and the evolution of the P3 could be accurately explained by a model including mutually inhibiting accumulators for the two categorical choice outcomes, and a context-dependent urgency signal that grows linearly over time. In a second experiment using a variation of the same paradigm, we replicated the finding that the P3 tracks categorical choice formation and we showed that it can be used to track changes of mind. We further show that the P3 tracks not only linear increases in urgency over time, but also sudden changes in the speed/accuracy regimes over the course of a single decision. These findings provide a valuable neural marker that reflects the interaction between external (evidence) and internal (urgency) sources of information during evidence accumulation.

The compulsive behaviour typically observed in patients with obsessive-compulsive disorder (OCD) is likely related to abnormalities in decision-making. The inability to commit to ultimate decisions, for example, patients unable to decide whether their hands are sufficiently clean, may reflect failures in accumulating sufficient evidence before a decision. I used a low-level perceptual discrimination task (the random-dot-motion task, RDMT) and two response conflict control tasks to investigate the process of evidence accumulation in OCD, specifically to test the hypothesis of enhanced evidence accumulation in OCD patients relative to healthy volunteers. I will show how healthy humans and OCD patients differ in their response strategies/decision-making processes and will provide evidence for impaired decision-formation processes in OCD, with a differential influence of high and low uncertainty contexts on evidence accumulation (decision threshold) and on the quality of evidence gathered (drift rates). I will further emphasise that OCD patients are sensitive to monetary incentives, heightening speed in the speed-accuracy tradeoff, improving evidence accumulation.

TBC