Differential contributions of the two human cerebral hemispheres to action timing

Abstract

Rhythmic actions benefit from synchronization with external events. Auditory-paced finger tapping studies indicate the two cerebral hemispheres preferentially control different rhythms. It is unclear whether left-lateralized processing of faster rhythms and right-lateralized processing of slower rhythms bases upon hemispheric timing differences that arise in the motor or sensory system or whether asymmetry results from lateralized sensorimotor interactions. We measured fMRI and MEG during symmetric finger tapping, in which fast tapping was defined as auditory-motor synchronization at 2.5 Hz. Slow tapping corresponded to tapping to every fourth auditory beat (0.625 Hz). We demonstrate that the left auditory cortex preferentially represents the relative fast rhythm in an amplitude modulation of low beta oscillations while the right auditory cortex additionally represents the internally generated slower rhythm. We show coupling of auditory-motor beta oscillations supports building a metric structure. Our findings reveal a strong contribution of sensory cortices to hemispheric specialization in action control.

Keywords: auditory cortex; beta partial directed coherence; finger tapping; hand motor control; human; lateralization; neuroscience; theta oscillations.

Figure 1.. Tapping conditions.

Participants were instructed to tap either to every beat (fast tapping, left panel) or to the fourth position of four consequent auditory beats (slow condition, right panel). Filled squares represent tapping events, white squares represent auditory beats without tapping in the slow tapping condition.

Figure 2.. Overview of the applied MEG analysis steps.

(1) Coherent sources with the EMG signal were detected at fast tapping frequency using a DICS beamformer. (2) Timeseries were extracted from the localized sources using an LCMV beamformer. (3) A sliding window time-frequency analysis was applied to transform these signals into a time-frequency-representation. By averaging over frequencies (14–20 Hz) a low beta band signal was extracted. (4) Source signals were fed into a time and frequency resolved directed connectivity analysis (TPDC).

Figure 3.. Asymmetries in performance.

Timing variability is defined as standard deviation of the absolute distance between the actual and target inter-tap-intervals (Pflug et al., 2017). Smaller values are associated with better performance. Error bars represent the within subject standard error of the mean. While the right hand taps more precisely in fast tapping, the left hand demonstrates lower timing variability in the slow tapping condition. The interaction between hand and condition is significant at p = 0.003. Note the overall higher precision in fast compared to slow tapping (differently scaled y-axes; Repp, 2005).

Figure 4.. Brain areas activated by rhythmic finger tapping.

Red: BOLD activation associated with slow tapping (p<0.05, FWE cluster-level corrected). Blue: BOLD activation associated with fast tapping (p<0.05, FWE cluster-level corrected). Yellow: Overlap of activity associated with slow and fast tapping. 3 and −6 indicate coronal and sagittal coordinates, respectively.

Figure 5.. Effects of internal generation of a slow rhythm.

BOLD activation for slow compared to fast tapping (p<0.05, FWE cluster-level corrected). Activity in the auditory association cortex is right-lateralized at p<0.001.

Figure 6.. Upper panels: Low beta band (14–20 Hz normalized to mean over conditions) power modulation in the left (upper left panel) and right (upper right panel) auditory association cortex (A2) for slow (red) and fast (blue) tapping.

One sequence of four auditory beats is illustrated. There was a stronger representation of the fast auditory beat frequency in the left compared to the right auditory association cortex during slow finger tapping (for statistics in the spectral domain, please see main text). Data are aligned to the tap at beat position four. Note the different scales for the beta power in left and right auditory association cortex. Shaded error bars represent the standard error of the mean (SEM). Lower panels: The background illustrates the low beta power in single slow tapping trials. Two sequences of four auditory beats with taps at beat position four are illustrated. Red curves represent mean low beta power ± SEM. Data are aligned to the right beat four in the panels.

Figure 7.. Upper left panel: Low beta band [14–20 Hz] power modulation in the supplementary motor area (SMA) for slow and fast tapping (mean over sequences).

During fast tapping (blue) the low beta power is modulated by the fast tapping rate while during slow tapping (red) there is a temporal modulation by the slow tapping rate (linear beta power decrease). Data are aligned to tap at beat position four. Shaded error bars indicate the standard error of the mean (SEM). For statistics in the spectral domain, please see main text. Upper right panel: The background illustrates the low beta power in single slow tapping trials. Two sequences of four auditory beats with taps at beat position four are illustrated. Data are aligned to the tap at the right beat position four in the panels. Same scale as in Figure 6. Red curves represent mean low beta power ± SEM. Lower panels: Time frequency representation of the SMA source signal during fast (left panel) and slow tapping (right panel).

Figure 8.. Differences in low beta power modulation between too short and too long inter-tap-intervals (ITI).

Sequences of four auditory beats with taps at beat four were aligned at the left tap in the first three panels. While the first three panels illustrate effects during slow tapping, the right panel illustrates low beta power in the supplementary motor area (SMA) during fast tapping (data left aligned). Significant differences between too long and too short sequences were marked in grey. Only low beta power in the right auditory association cortex (A2) and in the SMA predicted performance during slow tapping. Low beta amplitude coded the ITI during fast tapping in the SMA. Note the different scales in the panels.

Figure 9.. Time-resolved partial-directed coherence (TPDC) during slow (upper panels) and fast tapping (lower panels).

TPDC was particularly strong in the low beta and the low gamma band. Note the overall increased connectivity strength between the left auditory association cortex and the SMA compared to the other connections. SMA: supplementary motor area. A2: Auditory association cortex.

Figure 10.. Left panel: Condition differences between slow and fast tapping in low beta band [14–20 Hz] effective connectivity.

Right panel: Connections with increased low beta band interactions (p<0.05) during slow compared to fast tapping. SMA: supplementary motor area. A2: Auditory association cortex. TPDC: Time-resolved partia-directed coherence.