Although the basal ganglia play an important role in self-generated movement, their involvement in externally paced voluntary movement is less clear. We recorded local field potentials (LFPs) from the region of the subthalamic nuclei of eight patients with Parkinson's disease during the performance of a warned reaction time task in which an imperative cue instructed the subject to move or not to move. In 'go' trials, LFP activity in the beta frequency band ( approximately 20 Hz) decreased prior to movement, with an onset latency that strongly correlated with mean reaction time across patients. This was followed by a late post-movement increase in beta power. In contrast, in 'nogo' trials the beta power drop following imperative signals was prematurely terminated compared with go trials and reversed into an early beta power increase. These differences were manifest as power increases when go trials were subtracted from nogo trials. In six patients these relative beta power increases in nogo-go difference trials were of shorter latency than the respective reaction time. The findings suggest that, firstly, the subthalamic nucleus is involved in the preparation of externally paced voluntary movements in humans and, secondly, the degree of synchronization of subthalamic nucleus activity in the beta band may be an important determinant of whether motor programming and movement initiation is favoured or suppressed.

This is the unspecified version of the paper.This version of the publication may differ from the final published version. Permanent

This is the unspecified version of the paper.This version of the publication may differ from the final published version. average 72 and 139 ms respectively) to a numerical counter. The movement of the eyes was used to start the counter incrementing once every second, with the exception that the duration of the first number could be varied between 400 and 1600 ms. Subjects had to say whether the time they had seen the first digit was more or less than that for the subsequent digits (a constant 1 s). suggesting that the illusion of chronostasis is linked to the time taken to move the eyes. In fact, subjects appeared to extend the time that they thought they had seen the first target back in time to approximately 50 ms prior to the start of eye movement. Although subjects reported no awareness of the counter changing during their saccades, it is possible that they were able to use this digit shift as a cue to initiate time judgements. This would invalidate the matched times we calculated (measured from the moment the eyes actually reached the counter). However, a control experiment in which the counter was triggered either very early or very late during a large (55º) saccade showed no difference in the duration of chronostasis, despite modifying the period that the digit was actually on screen by 85 ms. Permanent repository link 3The tight coupling of the duration of chronostasis to the duration of the saccade suggests that the effect may be linked to the perceptual "gap" caused by saccadic suppression and retinal blur that occurs when we move the eyes 3,4 . However, it is possible that the illusion of chronostasis is not tightly coupled to movement of the eyes per se, but occurs because subjects also shift the locus of their visual attention at around the time their eyes move 5 . This attention shift may act as the reference point to which the target is predated. In order to test this, subjects were asked either to make the usual saccade to target or first to shift their attention to the target and then move their eyes. Figure 2a shows that the illusion of chronostasis persisted with a similar magnitude when subjects shifted their attention before moving their eyes. Control trials intermixed with the eye movement trials verified that subjects were successful in shifting the locus of their visual attention 6 . They fixated a central cross and had to saccade to a target appearing on the right or left of the screen. If they had been told to shift their attention to the correct side before the target appeared, their reaction time was faster than if they had been incorrectly cued (Fig 2b).Although chronostasis is linked to voluntary saccades, the coupling is not obligatory:there is at least one condition under which the illusion is not experienced. We designed a third experiment in which the positional stability of the target counter was systematically broken.Subjects made a saccade to target, but in some trials the computer displaced the target by up to 9 degrees during the time the eyes were moving. Under such condi...

This is the accepted version of the paper.This version of the publication may differ from the final published version. Permanent repository link

Activation of the basal ganglia has been shown during the preparation and execution of movement. However, the extent to which the activation during movement is related to efferent processes or feedback-related motor control remains unclear. We used motor imagery (MI), which eliminates peripheral feedback, to further investigate the role of the subthalamic area in the feedforward organization of movement. We recorded local field potential (LPF) activity from the region of the subthalamic nucleus (STN) in eight patients with Parkinson's disease off dopaminergic medication during performance of a warned reaction time task. Patients were instructed to either extend the wrist [motor execution (ME)], to imagine performing the same task without any overt movement (MI), or, in a subgroup, to perform a non-motor visual imagery (VI) task. MI led to event-related desynchronization (ERD) of oscillatory beta activity in the region of the STN in all patients that was similar in frequency, time course and degree to the ERD occurring during ME. The degree of ERD during MI correlated with the ERD in trials of ME and, like ME, was accompanied by a decrease in cortico-STN coherence, so that STN LFP activity during MI was similar to that in ME. The ERD in ME and MI were both significantly larger than the ERD in VI. In contrast, event-related synchronization (ERS) was significantly smaller in trials of MI, and even smaller in trials of VI, than during ME. The data suggest that the activity in the region of the human STN indexed by the ERD during movement is related to the feedforward organization of movement and is relatively independent of peripheral feedback. In contrast, sensorimotor feedback is an important factor in the ERS occurring in the STN area after completion of movement, consistent with a role for this region in trial-to-trial motor learning or the re-establishment of postural set following movements.

No abstract

Several lines of evidence suggest that motoric brain structures may form the core amodal component of a neural network supporting a wide range of timed behaviours. Here, we review recent findings which elucidate the neural computations that occur within motor regions, and in particular the supplementary motor area, in order to support precisely timed actions. Although motor activity may help us represent time, it is also clear that action both enriches and complicates the interpretation of sensory inputs. Hence, in the second half of this review, we also consider the latest findings regarding the perceptual distortions that our actions can impose upon the subjective timeline.3

No abstract