Temporal expectation task

The experimental paradigm consisted of one session assessing temporal expectation of human body segment versus inanimate object motion. This task has been described elsewhere [9]. Briefly, it consisted of two different perceptual tasks: a target task, involving the perception of a common movement of a human body segment (a right hand writing a sentence), and a control task, involving the perception of a movement of an inanimate object (a ball reaching a target). These movements were presented as video files on a computer screen at a fixed speed. The human body motion video displayed the movement of a hand writing a proverb sentence in Italian (‘il mattino ha l’oro in bocca’) with a black pencil on an A4-size white paper sheet positioned on the table; the inanimate object motion video consisted of a linear movement, directed from left to right, of a black ball on a white screen heading towards a black vertical bar. The two videos displayed movements under the same visual angle, had the same total duration (20 s), and the order of administration was counterbalanced across the two groups. The tasks have been programmed using dedicated software (E-Prime 2.0, SciencePlus). For each of the two tasks, subjects were shown the full video for a single time without receiving any instruction. The actual task consisted of watching the same video, but it was interrupted after a variable interval from its onset (pre-dark interval) by a dark interval of variable duration (dark interval). During the dark interval, subjects were asked to indicate when the movement represented in the video reached its end by clicking on the space bar of the keyboard.

Thus, participants used their knowledge of velocity of motion to extrapolate duration of the movement of the body segment (or object) represented in the video. For the sake of the analysis, we refer to the period spanning from the beginning of the dark interval to the subject’s response as ‘reproduced interval’, whereas we refer to the real duration of the dark interval as ‘target interval’. Participants did not receive any feedback on interval duration or their performance throughout the whole session. Three different dark intervals (6, 9 and 12 s) were used. Separately for each task (target and control task), each condition (6, 9 and 12 s dark intervals) was administered six times in a randomized order, for 36 trials per experimental session. Handwriting (target task) and inanimate object motion (control task) were presented in separate blocks; the order of presentation of the blocks was random across subjects. The experimental design is summarised in Fig. 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. A schematic view of the experimental design.

Healthy volunteers received one session of either active, inhibitory, 1Hz (Group1) or sham (Group2) repetitive transcranial magnetic stimulation (rTMS) covering the right lateral cerebellum prior the execution of a temporal expectation task. Two videos were shown once to the subjects: “writing” video showed a common movement of a human body segment (a right hand writing a sentence), while “ball” video showed a movement of an inanimate object (a ball reaching a target). The task consisted in watching the same video, but interrupted, after a variable interval from its onset by a dark interval of variable duration (6, 9 and 12 seconds). During the dark interval subjects were required to judge the duration of video clicking on space bar when they reckoned that the movement had reached its end.

https://doi.org/10.1371/journal.pone.0116607.g001

rTMS procedure

TMS was performed with a Magstim Rapid magnetic stimulator (Magstim Company, Whitland, Dyfed, UK) connected to a figure-of-eight coil (wing diameters of 90 mm) positioned over the right lateral cerebellum (3 cm right and 1 cm inferior to the inion). The coil was positioned tangentially to the scalp with the handle pointing up. The current in the coil was directed upwards, which induced a downward current in the cerebellar cortex [18]. The scalp coordinates were the same adopted in previous studies, in which MRI reconstruction and neuro-navigation systems showed that cerebellar TMS in this site predominantly target the posterior and superior lobules of the lateral cerebellum [15,16]. This approach has been adopted in previous investigations in which cerebellar rTMS was shown to be effective in modulating the excitability of the contralateral motor cortex [19,20] and interfering with cognitive functions such as procedural learning and sub-second time perception [21,22]. The exact coil position was marked by an inking pen to ensure an accurate positioning of the coil throughout the experiment. The stimulating coil was held by hand and coil position was continuously monitored throughout the experiment. Stimulation of structures deeper than the cortical cerebellar layers seems unlikely, owing to the fact that the effects of the magnetic field do not go beyond 2–3 cm below the scalp [18].

A single train (duration: 10 minutes; 600 stimuli) of 1Hz-rTMS was delivered at 90% of the motor threshold (MT) on right cerebellum. For sham rTMS the coil was positioned over the same scalp site, but angled away (90-degrees) so that no current was induced in the brain.

The MT was determined as the minimum stimulus intensity capable of eliciting a motor evoked potential (MEP) of at least 50 μV in the contralateral first dorsal interosseus (FDI) muscle in five or more of ten consecutive stimulations. EMG was recorded with silver disc surface electrodes placed in a tendon belly arrangement over the FDI muscle. The ground electrode was placed at the wrist. EMG signals were amplified and filtered (20 Hz to 1 kHz) with a D360 amplifier (Digitimer Limited, Welwyn Garden City, UK). The signals were sampled at 5000 Hz, digitized using a laboratory interface (Power1401, Cambridge Electronics Design, CED, Cambridge, UK) and stored on a personal computer for display and later off-line data analysis.

Data analysis

Performance on the temporal expectation task was analysed measuring the timing error, the normalized absolute value of timing error, the coefficient of variability and the percentage of anticipation responses. These parameters allowed us to describe the ability in time processing determining the magnitude (absolute value of timing error) and the variability (coefficient of variability) of the timing error, as well as the strategy adopted by subjects during time prediction (timing error and percentage of anticipation responses).

The timing error was defined as the time elapsing between the subject’s response and the real end of the video, and had a negative value when the response preceded the real end of the video, and positive when the response was delayed in respect to the real end of the video. The absolute value of this error was then normalized with respect to the corresponding target interval and was expressed as a percentage. This parameter provides a measure of the accuracy of subjects in estimating the corresponding target interval: the greater the absolute error, the greater the error in temporal judgement regardless of its direction.

The coefficient of variability was calculated as Standard Deviation (SD)/mean of the reproduced intervals, representing an index of performance variability in temporal judgement. Finally, the percentage of anticipation responses (calculated on the total number of responses) represents the number of responses in which subjects pressed the space button before the real end of the movement (negative error). Together with the timing error, this parameter gives information on the tendency to under- or over-estimate the target interval.

Statistical analysis

Timing parameters (timing error, absolute value of timing error, coefficient of variability and percentage of responses in anticipation) were analysed by means of a repeated measures ANOVA with group (1Hz-rTMS and Sham) as between subject factor and task (target, control) and target interval (6, 9 and 12 s) as within- subjects factors. Post hoc analysis of significant interactions was performed by means of t-tests applying the Bonferroni correction for multiple comparisons where necessary. P-values of < 0.05 were considered as threshold for statistical significance. Statistical analysis was performed with SPSS 13.0.

The role of cerebellum in temporal expectation

Temporal expectations make use of timing information in order to optimise motor or perceptual performance [8]. For example, to perceptually estimate whether a car might cross one’s way and to react in time, the sensorimotor system would benefit from predicting the car’s upcoming movement based on sensory input and constantly recalibrating sensory predictions to match the ever-changing properties of one’s sensorimotor system and environment. The same proof of concept can be applied in the temporal estimation of a human motor act. Let’s think about a tennis player, who has to predict the end of his opponent’s tennis shot in order to react in time.

In agreement with the role of cerebellum in processing and adapting motor related information, it has been shown that the cerebellum is needed to optimize self-action by recalibrating predictions capturing the sensory consequences of one’s actions [30–32]. However, the cerebellum seems also implicated in predictions exclusively related to a perceptual domain. Indeed, when subjects used temporal information inherent the spatial-temporal trajectory of a dynamic visual stimulus to predict its final position, activations were reported in left lateralized inferior parietal cortex [10], sensorimotor regions of premotor and parietal cortices [11], and cerebellum [12].

In the present study, we wanted to expand this topic by exploring the role of cerebellum in predictions related to a visually perceived motor act (here, handwriting). We found that subjects’ precision in in temporally predicting the end of a human movement (referred here to the absolute value of timing error) was influenced by 1Hz-rTMS disruption, whereas the variability in temporal expectation performance and the direction of the timing error (expressed by both the timing error and the percentage of responses in advance) were not influenced by 1Hz-rTMS disruption.

In self-action it is well known that cerebellum is involved in predictive control of movement [33]. The term “predictive” refers to the feed-forward portion of movement that is planned in advance and is unchanged by online peripheral feedback during motor execution. Forward models capture the expected sensory consequences of one’s actions based on current body state and the efference copies of motor commands [33] and enable fast and accurate movements despite delayed or missing sensory feedback [34–37].

Here, we demonstrated for the first time that cerebellum is involved in predictive movement control in absence of self-action; i.e., during observation of a motor act, thus confirming recent theories supporting a cognitive-perceptual role of the cerebellum [38]. In accordance, it has been suggested that the cerebellum is part of the human action observation network [39–42] mediating the interactions of self-executed movements in biological motion perception (i.e., action-perception coupling) [43].

The perceptual predictive movement control operated by the cerebellum might be implicated in motor planning and its disruption might contribute to some movement abnormalities exhibited by patients with cerebellar damage, like lack of coordination, and poor accuracy [44]. However to what extent the prediction of temporal outcome of observed motor actions operated by the cerebellum may affect motor execution has to be determined in future studies.

Intriguingly, when cerebellar activity was disrupted by inhibitory 1Hz-rTMS we did not observe any modification in time estimation when the perceived movement was that of a moving object (not a biological motion). Although this does not exclude a cerebellar involvement in predicting the outcome of a moving object [12], it nevertheless seems consistent with previous studies suggesting cerebellar activation during the encoding of a biological motion with respect to other stimuli [29,42,45]. This suggests that cerebellar structures are specifically involved in processing the temporal aspects of a movement belonging to the human repertoire thus probably making this task more sensible to rTMS disruption.

In addition, there are some methodological characteristics of the study conducted by O’Reilly and co-workers that could partially justify this discrepancy, like the presence of an occluder, the type of task (temporo-spatial instead of temporal only), the type of participants response (a qualitative dycotomic one instead of a quantitative one) [12].

Another noteworthy aspect of our results is that cerebellar disruption led to an impact on supra-second intervals. Most studies available in the literature suggest that the cerebellum is involved in sub-second timing, although some lesion studies suggest that the cerebellum is also involved in supra-second intervals [25]. Notably, two studies, using repetitive transcranial magnetic stimulation, determined that the cerebellum is essential in timing in the range of millisecond time intervals and not of supra-second time intervals [16,46]. However, these studies analysed the role of cerebellum in explicit motor and perceptual timing, whereas our study focused on temporal expectation, i.e., an implicit perceptual timing task, likely engaging a different cerebral network [47]. In addition, to our knowledge this is the first study that directly analyse the role of cerebellum in time processing of a motor act, whereas previous studies using rTMS adopted tasks requiring either motor production or perceptual discrimination of a timed duration [16,46].

It’s worthy to note that although we demonstrated that lateral cerebellum is directly involved in a temporal expectation task whose goal is to predict the temporal outcome of a motor act, we can only do some speculation about the cortical areas that may be recruited in this perceptual-motor ability. Indeed, given the cerebellar connectivity with cortical regions engaged in higher-order perceptual processing and spatiotemporal awareness (e.g., posterior parietal and premotor cortex) [48,49], we may speculate that a cerebral network comprising premotor cortex, parietal cortex and the cerebellum could guarantee an adequate timing prediction. However, the precise definition of the cerebral network engaged in everyday temporal expectations related to human motor acts needs to be elucidated in future studies.

One might argue that changes in motor performance, induced by cerebellar 1Hz-rTMS might have influenced the performance on the temporal expectation task. When applied to lateral cerebellum, 1Hz-rTMS is known to induce a disruption of cerebellar outputs to the contralateral motor cortex (M1). Low-frequency right cerebellar stimulation has been demonstrated to induce a decrease in the excitability of intracortical facilitatory circuits in left M1, likely due to the reduction of excitatory drive of cerebellar nuclei to the motor cortex, thus reflecting a suppressive effect of low-frequency rTMS on cerebellar Purkinje cells [20]. Further, low-frequency right cerebellar stimulation can influence motor performance [17]. Since our subjects were asked to press the space button with their right hand to make their responses, the observed modifications in the execution of the temporal prediction task might have been caused by a modification in the left M1 excitability and/or right hand performance. However, the impairment in temporal prediction selectively in the target task (handwriting) and not in the control one (moving ball) cannot be attributed to an effect of 1Hz-rTMS on motor function because identical motor responses were required for the two conditions.