Bayesian Integration and Non-Linear Feedback Control in a Full-Body Motor Task (original) (raw)
Related papers
Stevenson 2009 Bayes Int Feedback Fullbody
A large number of experiments have asked to what degree human reaching movements can be understood as being close to optimal in a statistical sense. However, little is known about whether these principles are relevant for other classes of movements. Here we analyzed movement in a task that is similar to surfing or snowboarding. Human subjects stand on a force plate that measures their center of pressure. This center of pressure affects the acceleration of a cursor that is displayed in a noisy fashion (as a cloud of dots) on a projection screen while the subject is incentivized to keep the cursor close to a fixed position. We find that salient aspects of observed behavior are well-described by optimal control models where a Bayesian estimation model (Kalman filter) is combined with an optimal controller (either a Linear-Quadratic-Regulator or Bang-bang controller). We find evidence that subjects integrate information over time taking into account uncertainty. However, behavior in this continuous steering task appears to be a highly non-linear function of the visual feedback. While the nervous system appears to implement Bayes-like mechanisms for a full-body, dynamic task, it may additionally take into account the specific costs and constraints of the task.
Bayesian decision theory in sensorimotor control
Trends in cognitive sciences, 2006
Action selection is a fundamental decision process for us, and depends on the state of both our body and the environment. Because signals in our sensory and motor systems are corrupted by variability or noise, the nervous system needs to estimate these states. To select an optimal action these state estimates need to be combined with knowledge of the potential costs or rewards of different action outcomes. We review recent studies that have investigated the mechanisms used by the nervous system to solve such estimation and decision problems, which show that human behaviour is close to that predicted by Bayesian Decision Theory. This theory defines optimal behaviour in a world characterized by uncertainty, and provides a coherent way of describing sensorimotor processes.
Probabilistic optimization in the human perceptuo-motor system
The Journal of Physical Fitness and Sports Medicine, 2013
Despite the variability of internal and external environments, the human central nervous system (CNS) can generate precise and stable perception and motor behaviors. What mechanism enables this ability? Answering this question is one of the significant goals in the human sciences, including neuroscience, cognitive science, physical education and sports science. The Bayesian integration theory proposes that the CNS learns the prior distribution of a task and integrates it with sensory information to minimize the effect of sensory noise. In this article, we introduce psychophysical reports using motor timing and temporal order judgment (TOJ) tasks that support the Bayesian integration theory. Subsequently, we demonstrate the event-related potentials (ERPs) behind Bayesian integration that operates in somatosensory TOJ.
Bayesian integration in sensorimotor learning
Nature, 2004
When we learn a new motor skill, such as playing an approaching tennis ball, both our sensors and the task possess variability. Our sensors provide imperfect information about the ball's velocity, so we can only estimate it. Combining information from multiple modalities can reduce the error in this estimate 1-4 . On a longer time scale, not all velocities are a priori equally probable, and over the course of a match there will be a probability distribution of velocities. According to bayesian theory 5,6 , an optimal estimate results from combining information about the distribution of velocities-the prior-with evidence from sensory feedback. As uncertainty increases, when playing in fog or at dusk, the system should increasingly rely on prior knowledge. To use a bayesian strategy, the brain would need to represent the prior distribution and the level of uncertainty in the sensory feedback. Here we control the statistical variations of a new sensorimotor task and manipulate the uncertainty of the sensory feedback. We show that subjects internally represent both the statistical distribution of the task and their sensory uncertainty, combining them in a manner consistent with a performance-optimizing bayesian process 4,5 . The central nervous system therefore employs probabilistic models during sensorimotor learning.
Optimality of human movement under natural variations of visual–motor uncertainty
Biological movements are prone to error. Different movements lead to different errors, and the distributions of errors depend on movement amplitude and direction. Movement planning would benefit from taking this variability into account, by applying appropriate corrections for movements associated with the different shapes and sizes of error distributions. Here we asked whether the human nervous system can do so. In a game-like task, participants performed rapid sequences of goal-directed pointing movements in different directions, toward stimulus configurations presented at different eccentricities on a slanted touch screen. The task was to accumulate rewards by hitting target regions and to minimize losses by avoiding penalty regions. The distributions of endpoint errors varied in size and degree of anisotropy across stimulus locations. Our participants adjusted their movements toward the different locations accordingly. We compared human behavior with the optimal behavior predicted by ideal movement planner maximizing expected gain. In most cases, human behavior was indistinguishable from optimal. This is evidence that human movement planning approaches statistical optimality by representing the task-relevant movement variability.
Journal of Neuroscience, 2012
Sensory-motor behavior results from a complex interaction of noisy sensory data with priors based on recent experience. By varying the stimulus form and contrast for the initiation of smooth pursuit eye movements in monkeys, we show that visual motion inputs compete with two independent priors: one prior biases eye speed toward zero; the other prior attracts eye direction according to the past several days' history of target directions. The priors bias the speed and direction of the initiation of pursuit for the weak sensory data provided by the motion of a low-contrast sine wave grating. However, the priors have relatively little effect on pursuit speed and direction when the visual stimulus arises from the coherent motion of a high-contrast patch of dots. For any given stimulus form, the mean and variance of eye speed covary in the initiation of pursuit, as expected for signal-dependent noise. This relationship suggests that pursuit implements a trade-off between movement accuracy and variation, reducing both when the sensory signals are noisy. The tradeoff is implemented as a competition of sensory data and priors that follows the rules of Bayesian estimation. Computer simulations show that the priors can be understood as direction-specific control of the strength of visual-motor transmission, and can be implemented in a neural-network model that makes testable predictions about the population response in the smooth eye movement region of the frontal eye fields.
Optimality in human motor performance: Ideal control of rapid aimed movements
Psychological Review, 1988
A stochastic optimized-submovement model is proposed for Pitts' law, the classic logarithmic tradeoff between the duration and spatial precision of rapid aimed movements. According to the model, an aimed movement toward a specified target region involves a primary submovement and an optional secondary corrective submovement. The submovements are assumed to be programmed such that they minimize average total movement time while maintaining a high frequency of target hits. The programming process achieves this minimization by optimally adjusting the average magnitudes and durations of noisy neuromotor force pulses used to generate the submovements. Numerous results from the literature on human motor performance may be explained in these terms. Two new experiments on rapid wrist rotations yield additional support for the stochastic optimizedsubmovement model. Experiment 1 revealed that the mean durations of primary submovements and of secondary submovements, not just average total movement times, conform to a square-root approximation of Pitts' law derived from the model. Also, the spatial endpoints of primary submovements have standard deviations that increase linearly with average primary-submovement velocity, and the average primary-submovement velocity influences the relative frequencies of secondary submovements, as predicted by the model. During Experiment 2, these results were replicated and extended under conditions in which subjects made movements without concurrent visual feedback. This replication suggests that submovement optimization may be a pervasive property of movement production. The present conceptual framework provides insights into principles of motor performance, and it links the study of physical action to research on sensation, perception, and cognition, where psychologists have been concerned for some time about the degree to which mental processes incorporate rational and normative rules.
Visual Abstract Recent work suggests that the brain represents probability distributions and performs Bayesian integration during sensorimotor learning. However, our understanding of the neural representation of this learning remains limited. To begin to address this, we performed two experiments. In the first experiment, we replicated the key behavioral findings of Körding and Wolpert (2004), demonstrating that humans can perform in a Bayes-optimal manner by combining information about their own sensory uncertainty and a statistical distribution of lateral shifts encoun-Significance Statement Generalization provides unique insights into the motor learning process. However, this type of learning has typically been investigated using fixed or deterministic perturbations and noise-free feedback information, which are not naturalistic. Here, we replicate important findings indicating that information is integrated in a Bayes-optimal manner during sensorimotor learning under uncertainty. We then extend these findings by showing that this learning generalizes to the opposite limb. These results have implications for our understanding of the neural mechanisms of motor learning as well as practical application to the contexts of sport training and motor rehabilitation.