Bridging the neural and algorithmic correlates of action-stopping

Abstract

AbstractRapid stopping of actions is crucial for survival. The algorithm underlying action-stopping is often described using the independent race model, where the race between independent go- and stop-processes determines whether one is able to stop. However, it is unclear whether humans have neural correlates of these processes. Here, using scalp potentials from three studies, we show that the mu (8-13 Hz) desynchronization over the contralateral motor cortex follows the predictions of a go-process: the activity builds up till an action is made. We also show that the low beta (13-20 Hz) activity over the frontal areas follows the predictions of a stop-process: the activity builds up after the stop signal and prior to the stopping latency. Using these neural correlates, we observed that: 1) The failure to stop mostly results from a rapid go-process, 2) The go-process seems to be influenced by the stop-process during the late stages of action-stopping, 3) Intrinsic prefrontal beta activity has a retardive effect on movement, 4) Violations of independence assumption in the failed stop trials may result from the activity of the stop-process. Future studies may use these neural correlates to resolve the mimicry between models of action-stopping.