Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Distinct frequency channels for motor and sensory processing in speech production
Poster E71 in Poster Session E, Thursday, October 26, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Francesco Mantegna1, Joan Orpella2, David Poeppel1,3; 1Department of Psychology, New York University, New York, NY, USA., 2Department of Communicative Sciences and Disorders, New York University Steinhardt, New York, NY, USA, 3Ernst Strüngmann Institute for Neuroscience, Frankfurt, Germany
Speech production involves the interaction between action and perception. The successful integration between motor output and sensory input ensures the fluency that we typically experience during speech production. Motor and sensory systems deal with information having different physical properties (i.e., muscle contraction and air vibration). Different sources of information undergo a series of transformations at the brain level. Following these transformations does the brain operate in a common space or in separate spaces to integrate motor and sensory information? In other words, does sensorimotor integration involve conjoined or separated processing in motor and sensory cortices? Given the fast temporal rates that characterize speech production and perception, a fast neural mechanism is necessary for sensorimotor integration. One candidate mechanism to underwrite fast temporal aspects is neural oscillations. Frequency specific power modulations have been associated with both motor and sensory processing. Previous studies show that power desynchronization in alpha (8-12 Hz) and beta (13-30 Hz) frequency bands - collectively referred to as motor unit, or mu rhythms - is observed during motor execution. Mu rhythms are generated not only during overt but also during covert motor execution. If mu rhythms were generated only during overt motor execution, they could be a by-product of the different physical properties pertaining to input and output. Instead, the fact that mu rhythms are generated in the absence of movement and its sensory consequences suggests that they reflect neural computations independent from movement outcome. Mu rhythms are generated during a variety of motor tasks including limb and speech movements. Regardless of the movement effectors, mu rhythm generation has been localized to the motor and sensory cortices. Previous studies have grouped alpha and beta frequency bands together under the mu rhythm umbrella. However, this grouping may be misleading and fail to recognize the multifaceted nature of both alpha and beta frequency bands. In this study, we investigate the extent to which alpha and beta frequencies perform different functional operations during speech production. To this end, we used magnetoencephalography (MEG) to assess alpha and beta frequency band segregation in the time-frequency and spatial domain. Participants produced syllables (presented visually) either covertly (i.e., silently, imagined speech production, N=40) or overtly (i.e., aloud, N=40). Time-frequency analyses show robust alpha and beta desynchronization at the time of covert and overt speech production. Cluster-based permutation analysis across subjects shows a segregation of alpha and beta frequency bands into two different time-frequency clusters. The two clusters differ in timing: the beta cluster (200-300 ms) precedes the alpha cluster (400-500 ms). MEG source reconstruction shows that the beta cluster is localized in motor areas while the alpha cluster is localized in temporal areas. The results suggest that beta is primarily involved in motor planning and execution while alpha is primarily involved in sensory processing. We discuss the implications of having two separate frequency channels for sensorimotor integration in speech motor control.
Topic Areas: Speech Motor Control,