Visual modulation of proprioceptive reflexes during movement

Abstract

Previous research has demonstrated that feedback circuits such as reflexes can be tuned by setting their gains prior to movement onset during both posture and movement tasks. However, such a control strategy requires that perturbation contingencies be predicted during movement planning and that task goals remain fixed. Here we test the hypothesis that feedforward regulation of reflex circuits also occurs during the course of movement in response to changes in task goals. Participants reached to a visual target that was occasionally jumped on movement initiation, thus changing task goals. Reflex responses were elicited through a mechanical perturbation on the same trial, 100 ms after the target jump. Impedance to the perturbation was tuned to the direction of the preceding jump: reflex responses increased or decreased depending on whether the perturbation opposed or was consistent with the target jump. This modulation, although sensitive to the direction of the jump, was insensitive to jump amplitude, as tested in a follow-up experiment. Our findings thus suggest that modulation of reflex circuits occurs online, and is sensitive to changes in visual target information. In addition, our results suggest a two-level model for visuo-motor control that reflects hierarchical neural organization.

Introduction

In order to explain how the nervous system selects control signals that are appropriate for a desired movement, conventional models of control have proposed that movements are planned and executed in a sequential order. These models have often proposed that movement planning includes an optimization process, wherein kinematic or kinetic cost functions are minimized. Candidate cost functions have included mean squared jerk (Flash and Hogan, 1985), mean squared torque change (Uno et al., 1989), peak work (Soechting et al., 1995), or muscle energy (Goble et al., 2007) among others. Such optimization formulations often require that the entire movement trajectory be computed in advance, reinforcing the idea that biological control incorporates advance planning of movement trajectories. Support for this idea has come from several studies suggesting that the central nervous system accounts for limb and task dynamics when adapting to novel force conditions imposed by robotic manipulanda (Shadmehr and Mussa-Ivaldi, 1994, Goodbody and Wolpert, 1998), artificial gravity environments (Lackner and DiZio, 2003), or novel inertial loads (Sainburg et al., 1999). The prediction of such forces can be observed as “after-effects” following adaptation; if the imposed dynamics are removed after adaptation, movement errors emerge that mirror image the previously imposed forces, indicating anticipatory control.

According to this primarily feedforward control scheme, the role that sensory feedback plays during the course of movement is largely to correct deviations from the planned or reference trajectory. Such deviations to the movement path are caused by perturbations arising not only from the environment (Colgate and Hogan, 1988, Lackner and DiZio, 1994, Scheidt et al., 2001) but also from within the neuro-musculoskeletal system (Harris and Wolpert, 1998, Sainburg et al., 1999). When such perturbations can be predicted prior to movement onset, their effects can be partially accounted for during motion planning itself. In such cases, compensation for perturbations has been shown to occur through feedforward, reflex-mediated modulation of limb stiffness and viscosity, i.e. active components of limb impedance (Hore et al., 1990, Kimura et al., 2006, Lacquaniti and Maioli, 1989, Wang et al., 2001). Thus, resistance to perturbations has been shown to be selectively tuned based on pre-movement expectations of dynamic events. In addition, when the effects of perturbing forces can be learned over repeated exposure, and more importantly, when task goals remain constant, limb impedance has been shown to vary with the expected timing, and direction of the perturbing forces (Burdet et al., 2001, Franklin et al., 2003, Kimura et al., 2006, Wang et al., 2001). An elegant example of such anticipatory impedance modulation was provided by Lacquaniti et al. who, in a ball catching task demonstrated not only modulation, but also reversal of the reflex response to ball impact (Lacquaniti et al., 1991).

During everyday tasks however, environmental forces and perturbations can rarely be predicted prior to movement. For example, when reaching in a moving vehicle, unexpected disturbances arising from the environment can produce substantial forces on the body. In addition, task goals can often change during the course of a given movement. For instance, we frequently make movements toward objects that move en route, such as when reaching for a baby's hand or chasing a moving pet. We hypothesize that in order to effectively impede such perturbations, sensory feedback circuits are continuously modulated during the course of movement. We therefore predict that during reaching movements, modification of an ongoing movement in response to changes in visual information will also result in modulation of reflex responses to environmental perturbations. In a series of two experiments, using a novel experimental paradigm, we tested these predictions and also assessed the sensitivity of such modulation to changes in goal parameters. Our results confirmed that changes visual information, implemented through target displacements during the course of movement, resulted in significant modulation in reflex responses. Remarkably, our findings provided the first confirmation that short latency reflex responses were modulated during goal directed movement, independent of changes in background muscle activation or muscle state. Modulation of the longer latency responses were also consistent with changes in short latency reflex responses. This reflex modulation was sensitive to changes in target direction but surprisingly, insensitive to changes in target distance. This may reflect a fundamental limitation in online modulation of impedance through reflex mechanisms. It should also be emphasized that the observed changes in reflex responses occurred well before responses to changes in target location were initiated. Our results thus provide a demonstration of continuous feedforward tuning of sensory feedback during voluntary movement in response to changes in task goals.