How the Medial Premotor Cortex Encodes Elapsed Time
Hugo Merchant and Bruno B. Averbeck
(see pages 4552–4564)
Most people can easily tap along with an ongoing beat and continue tapping at the same frequency when external cues stop. This ability—particularly the ability to continue tapping in the absence of cues—depends on neural activity in the medial premotor cortex, which is thought to encode elapsed time between taps. How neurons represent timing information remains unknown, but several computational models have been proposed. These include drift-diffusion models, in which information accumulates over time until a threshold is reached, at which point an action is initiated. According to these models, differences in response times can occur if either the accumulation rate or the threshold changes.
To determine whether a drift-diffusion model might be used to encode elapsed time in the premotor cortex, Merchant and Averbeck trained monkeys to perform a synchronization-continuation task, in which animals tapped in synchrony with a cue and then continued to tap after the cue stopped. They found that the mean and standard deviation of the time intervals between monkeys' taps, as well as the distribution and other statistical features of the intervals, were consistent with the predictions of a drift-diffusion model.
The authors then recorded neural activity in the medial premotor cortex as monkeys performed the synchronization-continuation task, and used a decoding algorithm to estimate elapsed time between taps using population activity. Estimated elapsed times were generally consistent with actual times; but at the beginning and end of trials, the decoded times were sometimes shifted earlier or later than the actual time, suggesting the time counter did not always reset precisely when a tap occurred. Notably, during the continuation phase of the task (when no external timing cue was present), if neural activity suggested more time had elapsed then actually had, the monkeys tended to tap prematurely, whereas when the decoder lagged actual time, the tap was late. These results are consistent with the drift-diffusion model and suggest that population activity in the medial premotor cortex keeps track of elapsed time between actions in a rhythmic tapping task. With each tap, the internal timing mechanism resets, but resetting of the clock sometimes occurs somewhat before or after the tap occurs. Future work should investigate how the premotor cortex interacts with other components of timing circuitry to represent time across behaviors.
The ability to tap in rhythm depends on activity in the medial premotor cortex.
See Merchant and Averbeck for details.
The Computational and Neural Basis of Rhythmic Timing in Medial Premotor Cortex
Hugo Merchant and Bruno B. Averbeck
Journal of Neuroscience 23 March 2017, 37 (17) 4552-4564
The neural underpinnings of rhythmic behavior, including music and dance, have been studied using the synchronization-continuation task (SCT), where subjects initially tap in synchrony with an isochronous metronome and then keep tapping at a similar rate via an internal beat mechanism. Here, we provide behavioral and neural evidence that supports a resetting drift-diffusion model (DDM) during SCT. Behaviorally, we show the model replicates the linear relation between the mean and standard-deviation of the intervals produced by monkeys in SCT. We then show that neural populations in the medial premotor cortex (MPC) contain an accurate trial-by-trial representation of elapsed-time between taps. Interestingly, the autocorrelation structure of the elapsed-time representation is consistent with a DDM. These results indicate that MPC has an orderly representation of time with features characteristic of concatenated DDMs and that this population signal can be used to orchestrate the rhythmic structure of the internally timed elements of SCT.
How Monkeys Keep a Beat
Musicians use a metronome to set the pace of a piece of music. Without the metronome, the musician must rely on an internal beat-keeping mechanism to play music in time. Brain imaging studies have shown that themedial premotor cortex (MPC) is more strongly activated when an external timing cue (like a metronome) is no longer available, suggesting that this region is important for representing the amount of time between beats.
Researchers recorded the activity of single neurons in this region in two macaques as they tapped a button seven times at various instructed beats, the first four times with a visual or auditory cue and the next three times without a cue. The researchers found that MPC cells accurately represented the time between taps and the monkeys’ tapping behavior was consistent with a computational model of rhythmic timing that suggests information accumulates to a certain point and then triggers an action — in this case, a tap. Each tap then resets the internal timing mechanism.