Default mode network activity is related to efficiency in a combined motion error and gambling task.

Decision-making under risk lies at the core of strategic action in human activities. Nevertheless, the neural mechanisms underlying decision-making in trial-and-error-based learning in the absence of explicit risk ratios have not been thoroughly understood. This study investigates the neural correlates of long-term success during repeated risky decision-making. We combined a motion detection task with a higher-level risk task in a functional magnetic resonance imaging (fMRI) paradigm. Twenty-five healthy adults participated in an fMRI task that required stopping a ball covertly moving at a constant speed at a certain spot. Participants initially selected a task difficulty level progressively related to the magnitude of the monetary reward and then attempted the task of chosen difficulty. Positive versus negative feedback (a monetary gain versus no gain) contrast comprised clusters in the striatum, executive control regions, and the Default mode network (DMN). The activity in brain regions associated with metacognitive functions (DMN nodes) and not in those involved in reward evaluation and risk-taking (striatum) was positively related to the exploration index, which reflects the effectiveness of risk-taking. This study provides insights into the functional roles of the DMN (especially, its anterior node) in high-level decision-making in risky environments, highlighting its importance in metacognitive control during reward-related learning.