6533b820fe1ef96bd127aa18

RESEARCH PRODUCT

Motor decision and modular control of an hyper-redundant system

subject

description

This thesis is aimed at better understanding how the Central Nervous System (CNS) plans and controls voluntary movements. When moving, humans must overcome intrinsic (e.g. choosing which muscles to activate) and extrinsic (e.g. choosing where to reach an object) redundancy, requiring selecting one motor solution among several potential ones. To better understand this process, we studied in parallel two important motor control theories: muscular synergies and motor decision. In a first part, we focused on intrinsic redundancy by testing the muscular synergies hypothesis. According to it, the CNS simplifies the control of muscles, in using a limited set of building blocks whose linear combinations allow the performance of virtually any motor task. In this study, we challenge the modular motor control hypothesis by combining a) the design of a highly comprehensive experiment with b) the use of a unifying modularity model to describe single-trial EMG activity in space and time and c) a module evaluation process that assesses the modular decomposition not only in input space (EMG data reconstruction) but also in task space (task discrimination). Our rationale is that an effective modular control implementation would allow not only the formation of a wide variety of muscle patterns but also the achievement of a large set of tasks. The main theoretical result is the existence of few spatial and temporal modules that not onlygive a concise representation of muscle patterns but also carry nearly all task-relevant information of EMG signals. In a second part, we studied the decisional process that underlies all voluntary movement. In daily life, human movement is guided by objective external constraints (e.g. an object to reach), potential external cost/benefits (e.g. monetary reward) and internal cost/benefits associated with each movement (e.g. energy expenditure). Here,we aimed at investigating internal variables orienting action selection when facing the complexity of human-environment interactions. To this aim, we designed an experimental protocol reducing external constraints: no predetermined endpoint (e.g. salient target) and no explicit reward (e.g. money). Subjects had to perform whole body reaching movements towards a uniform surface (no pre-determined endpoint). Our results illustrate the presence of idiosyncratic values guiding posture and movement coordination that can be combined in a flexible manner as a function of context and subject. A first value takes into account the energy expenditure and articular jerk, while the other favored stable dynamic equilibrium but requires larger energy expenditure and articular jerk. In conclusion of this work, we suggest that motor control can be viewed as a decision process evaluating internal values to elaborate the most efficient control in function of context. In addition, this control can be simplified by the use of functional modules allowing CNS to generate rapidly a large set of whole body movements.