hide
Free keywords:
-
Abstract:
Human dexterity exceeds that of modern robots, despite omnipresent noise and vastly slower neural processing. Several lines of research have accrued evidence that the complex biological system is hierarchically organized, with building blocks or motor primitives under only partial cortical supervision. This study continues research on the hypothesis that viable candidates for such motor primitives are discrete and rhythmic movements, defined as point and limit cycle attractors, generated in the neuromechanical network. The distinction between rhythmic and discrete primitives was supported by a previous neuroimaging study that showed strikingly different cortical and subcortical activation: while rhythmic movements were associated with mostly primary motor areas, contralateral M1 and ipsilateral cerebellum, discrete movements involved a significantly broader network of cortical areas, including bilateral parietal and prefrontal regions. Given the rapid advances in imaging technology and analysis, the present study aims to replicate and extend these results from more than 10 years ago. Using the same movements, we recorded fMRI data with simultaneous recording of behavior. Subjects performed flexions and extensions with their right dominant wrist during acquisition of whole-brain scans in a 3T scanner; kinematic data were acquired by a custom-made goniometric device. In Experiment 1, movements were self-paced, performed in continuously rhythmic fashion or as single flexions and extensions self-initiated at random intervals over the 36-sec run. Using a GLM analysis, results largely replicated those of the previous study: rhythmic movement primarily elicited contralateral motor cortical, supplemental motor cortical, and ipsilateral cerebellar activations, whereas the discrete condition implicated a broader network of parietal and prefrontal areas, in addition to primary motor areas. In order to probe whether self-initiated timing was responsible for the extensive activations, Experiment 2 compared rhythmic and discrete movements that were visually cued and the number of initiations and terminations were matched. Preliminary results were largely consistent with the earlier study, but also raised additional questions. To shed light on the functional meaning of the activated network, novel connectivity analyses will be conducted. In addition, more focused follow-up measurements will be done using 9.4T scanning that provide significant increases in spatial resolution. These studies will provide a first important replication of previous influential results, supporting that even non-visually guided discrete movements require an extensiveset of cortical areas, while continuous rhythmic movements are generated with significantly less cortical substrate, but possibly rely on lower brainstem activation. This different neurophysiological substrate underscores that these two movement types actas different building blocks at different levels of the neural axis.