EP 106. Contributions of alpha and beta oscillations to bimanual sequence tapping

Abstract

Objective In 1:n bi-manual tapping studies participants intuitively choose their right hand for tapping the fast frequency. This preference is expressed in faster and more accurate performance compared to the opposite frequency mapping. This could be either due to a dominant role of the left hemisphere in planning and integration of movements ( Serrien et al., 2003 ) or due to a hemispheric specialization for relative frequencies [double filtering by frequency theory ( Robertson and Ivry, 2000 )]. In this framework, which has been proposed for sensory processing, the left hemisphere is optimized in processing higher relative frequencies while the right hemisphere excels in processing lower relative frequencies of a complex stimulus. We hypothesized comparable lateralization effects relative to movement frequencies. Since the strong tendency of the left hemisphere in controlling bi-manual learned sequences [pre-defined motor programs] would overlay a right hemisphere preference for lower frequencies, we expected to disclose frequency-related lateralization effects only in untrained sequences. Methods In 1:4 tapping paradigms, participants usually tap the first tone with both hands, continue tapping to every auditory tone with the fast hand, count until four and tap again with both hands. In our paradigm, participants were asked to wait for the fourth tone for the slow hand tapping, creating an unintuitive quadruple meter. We recorded magnetoencephalography (MEG) signals in 10 healthy right-handed subjects while performing optimal (left hand slow and right hand fast tapping) and non-optimal (opposite arrangement) bi-manual tapping conditions. As a control, uni-manual fast and slow conditions were performed with each hand alone and with both hands together. We reconstructed the timing signal in the alpha and beta bands of the underlying sources with beamforming. For comparing modulation of intra- and interhemispheric connectivity at different levels of the motor cortex hierarchy in optimal and non-optimal hand arrangements we used dynamic causal modeling. Results We found that the right hand benefit disappeared in all fast tapping conditions. Additionally, a left hand benefit for slow taps was found in the optimal bi-manual condition, which supports frequency-dependent lateralization effects. Comparing power spectra of the different positions in the sequence of four tones revealed two different effects in slow conditions. First and last position coincided with reduced alpha power while successive positions seem to be coded by a decrease in beta power. Changes in alpha power were found in right sensors, changes in beta power were found contralateral to the hand used for slow tapping. Conclusion Behavioral as well as spectral results fit the double filtering by frequency hypothesis. We therefore expected to find stronger intra-hemispheric connections in optimal hand arrangement indicating a local processing of information, whereas in non-optimal hand arrangement inter-hemispheric communication should be enhanced pointing to a higher amount of information exchange between the hemispheres.