Sensory uncertainty impacts avoidance during spatial decisions

Prior uncertainty impedes discrete locomotor adaptation

The impact of environmental uncertainty on locomotor adaptation remains unclear. Environmental uncertainty could either aid locomotor adaptation by prompting protective control strategies that stabilize movements to assist learning or impede adaptation by reducing error sensitivity and fostering hesitance to pursue corrective movements. To explore this, we investigated participants' adaptation to a consistent force field after experiencing environmental uncertainty in the form of unpredictable balance perturbations. We compared the performance of this group (Perturbation) to the adaptive performance of a group that did not experience any unpredictable perturbations (Non-Perturbation). Perturbations were delivered using a cable-driven robotic device applying lateral forces to the pelvis. We assessed whole-body center of mass (COM) trajectory (COM signed deviation), anticipatory postural adjustments (COM lateral offset), and first step width. The Perturbation group exhibited larger disruptions in COM trajectory (greater COM signed deviations) than the Non-Perturbation group when first walking in the force field. While the COM signed deviations of both groups decreased towards baseline values, only the Non-Perturbation group returned to baseline levels. The Perturbation groups COM signed deviations remained higher, indicating they failed to fully adapt to the force field before the end. The Perturbation group also did not adapt their COM lateral offset to counter the predictable effects of the force field as the Non-Perturbation group did, and their first step width increased more slowly. Our findings suggest that exposure to unpredictable perturbations impeded future sensorimotor adaptations to consistent perturbations.

Novel applications of Bayesian inference clarify sensorimotor uncertainty during stepping movements

The human nervous system relies on sensory information from the feet and legs to control the way we balance and walk. However, even in healthy individuals this sensory information is inherently variable and clouded with uncertainty. Researchers have found that the central nervous system (CNS) estimates body position amid the uncertainty of sensory signals in a way consistent with Bayesian inference. Bayesian inference posits that the brain accounts for variability in sensory data by combining it with learned expectations built from previous movement attempts. While initial findings on this topic are promising, they have neglected to study full-body movements such as gait and balance. The purpose of this research was to determine if the CNS controls balance-related stepping tasks in a way that fits a Bayesian framework. To address this purpose, we created a virtual reality protocol where participants moved their center of mass (CoM) to various targets while relying on uncertain visual cues and compensating for an alternating shift to the cursor position. We showed that as incoming sensory information became less certain, participants relied more on their learned expectation of body position and demonstrated more uncertainty in their responses. Accordingly, as participants learned to control and estimate their CoM position during our mobility task, they relied both on the sensory information they were receiving as well as learned expectations for its location. These results provide further evidence that the CNS is aware of the variability in sensory information and is proficient at compensating for the resultant uncertainty. We aim to apply these findings as a method for measuring the efficacy of interventions targeting sensory function.

Bayesian regression explains how human participants handle parameter uncertainty

Accumulating evidence indicates that the human brain copes with sensory uncertainty in accordance with Bayes’ rule. However, it is unknown how humans make predictions when the generative model of the task at hand is described by uncertain parameters. Here, we tested whether and how humans take parameter uncertainty into account in a regression task. Participants extrapolated a parabola from a limited number of noisy points, shown on a computer screen. The quadratic parameter was drawn from a bimodal prior distribution. We tested whether human observers take full advantage of the given information, including the likelihood of the quadratic parameter value given the observed points and the quadratic parameter’s prior distribution. We compared human performance with Bayesian regression, which is the (Bayes) optimal solution to this problem, and three sub-optimal models, which are simpler to compute. Our results show that, under our specific experimental conditions, humans behave in a way that is consistent with Bayesian regression. Moreover, our results support the hypothesis that humans generate responses in a manner consistent with probability matching rather than Bayesian decision theory.

How do humans make prediction when the critical factor that influences the quality of the prediction is hidden? Here, we address this question by conducting a simple psychophysical experiment in which participants had to extrapolate a parabola with an unknown quadratic parameter. We show that in this task, humans perform in a manner consistent with the mathematically optimal model, i.e., Bayesian regression.