Spatiotemporal factors influence sound-source segregation in localization behavior

To program a goal-directed response in the presence of acoustic reflections, the audio-motor system should suppress the detection of time-delayed sources. We examined the effects of spatial separation and inter-stimulus delay on the ability of human listeners to localize a pair of broadband sounds in the horizontal plane. Participants indicated how many sounds were heard and where these were perceived by making one or two head-orienting localization responses. Results suggest that perceptual fusion of the two sounds depends on delay and spatial separation. Leading and lagging stimuli in close spatial proximity required longer stimulus delays to be perceptually separated than those further apart. Whenever participants heard one sound, their localization responses for synchronous sounds were oriented to a weighted average of both source locations. For short delays, responses were directed towards the leading stimulus location. Increasing spatial separation enhanced this effect. For longer delays, responses were again directed towards a weighted average. When participants perceived two sounds, the first and the second response were directed to either of the leading and lagging source locations. Perceived locations were interchanged often in their temporal order (in ~40% of trials). We show that the percept of two sounds occurring requires sufficient spatiotemporal separation, after which localization can be performed with high accuracy. We propose that the percept of temporal order of two concurrent sounds results from a different process than localization, and discuss how dynamic lateral excitatory-inhibitory interactions within a spatial sensorimotor map could explain the findings.

Early stages of sensorimotor map acquisition: neurochemical signature in primary motor cortex and its relation to functional connectivity

Learning the mapping between movements and their sensory effects is a necessary step in the early stages of sensorimotor learning. There is evidence showing which brain areas are involved in early motor learning, but their role remains uncertain. Here, we show that GABA, a neurotransmitter linked to inhibitory processing, rises during and after learning and is involved in ongoing changes in resting-state networks.

“Characteristics and stability of sensorimotor activity driven by isolated-muscle group activation in a human with tetraplegia”

ABSTRACTThe topography and temporal stability of movement representations in sensorimotor cortex underlie the quality and durability of neural decoders for brain machine interface (BMI) technology. While single- and multi-unit activity (SUA and MUA) in sensorimotor cortex has been used to characterize the layout of the sensorimotor map, quantifying its stability has not been done outside of injury or targeted interventions. Here we aimed to characterize 1) the bilateral sensorimotor body map associated to isolated muscle group contractions and 2) the stability of multiunit firing responses for a single muscle (the extensor carpi radialis, ECR) over short (minutes) and long (days) time intervals. We concurrently recorded surface electromyograms (EMG) and MUA in a participant with incomplete high-spinal-cord injury as he executed (or attempted to execute) different metronome-paced, isolated muscle group contractions. Furthermore, for 8 recording sessions over 2 months, we characterized the sensorimotor map associated to ECR motions both within and across sessions. For each measurement period, we compared the stability of somatotopy (defined by the number of the channels on which a response was consistently detected) and firing pattern stability for each responsive channel. Stability was calculated for each channel in peri-EMG or peri-cue windows using both mean MUA firing rates and the full time-varying responses (i.e., MUA “shape”). First, we found that cortical representations of isolated group muscle contractions overlapped, even for muscles from disparate body regions such as facial and distal leg muscles; this was the case for both intact and de-efferented muscles, in both motor and sensory channels. Second, the spatial stability of somatotopy significantly changed over the course of both minutes and days, with the consistency between sessions decreasing across longer bouts of time. Firing pattern stabilities showed distinct profiles; mean MUA firing rates became less stable over time whereas MUA shape remained consistent. Interestingly, sensory channels were overall more consistent than motor channels in terms of spatial stability, mean MUA firing rates, and MUA shape. Our findings suggest that the encoding of muscle-driven specific activity in sensorimotor cortex at the level of MUA is redundant and widespread with complex spatial and temporal characteristics. These findings extend our understanding of how sensorimotor cortex represents movements, which could be leveraged for the design of non-traditional BMI approaches.

The Effect of Visual Uncertainty on Implicit Motor Adaptation

ABSTRACTSensorimotor adaptation is driven by sensory prediction errors, the difference between the predicted and actual feedback. When the position of the feedback is made uncertain, adaptation is attenuated. This effect, in the context of optimal sensory integration models, has been attributed to a weakening of the error signal driving adaptation. Here we consider an alternative hypothesis, namely that uncertainty alters the perceived location of the feedback. We present two visuomotor adaptation experiments to compare these hypotheses, varying the size and uncertainty of a visual error signal. Uncertainty attenuated learning when the error size was small but had no effect when the error size was large. This pattern of results favors the hypothesis that uncertainty does not impact the strength of the error signal, but rather, leads to mis-localization of the error. We formalize these ideas to offer a novel perspective on the effect of visual uncertainty on implicit sensorimotor adaptation.SIGNIFICANCE STATEMENTCurrent models of sensorimotor adaptation assume that the rate of learning will be related to properties of the error signal (e.g., size, consistency, relevance). Recent evidence has challenged this view, pointing to a rigid, modular system, one that automatically recalibrates the sensorimotor map in response to movement errors, with minimal constraint. In light of these developments, this study revisits the influence of feedback uncertainty on sensorimotor adaptation. Adaptation was attenuated in response to a noisy feedback signal, but the effect was only manifest for small errors and not for large errors. This interaction suggests that uncertainty does not weaken the error signal. Rather, it may influence the perceived location of the feedback and thus the change in the sensorimotor map induced by that error. These ideas are formalized to show how the motor system remains exquisitely calibrated, even if adaptation is largely insensitive to the statistics of error signals.

Early stages of sensorimotor map acquisition: learning with free exploration, without active movement or global structure

One of the puzzles of learning to talk or play a musical instrument is how we learn which movement produces a particular sound: an audiomotor map. The initial stages of map acquisition can be studied by having participants learn arm movements to auditory targets. The key question is what mechanism drives this early learning. Three learning processes from previous literature were tested: map learning may rely on active motor outflow (target), on error correction, and on the correspondence between sensory and motor distances (i.e., that similar movements map to similar sounds). Alternatively, we hypothesized that map learning can proceed without these. Participants made movements that were mapped to sounds in a number of different conditions that each precluded one of the potential learning processes. We tested whether map learning relies on assumptions about topological continuity by exposing participants to a permuted map that did not preserve distances in auditory and motor space. Further groups were tested who passively experienced the targets, kinematic trajectories produced by a robot arm, and auditory feedback as a yoked active participant (hence without active motor outflow). Another group made movements without receiving targets (thus without experiencing errors). In each case we observed substantial learning, therefore none of the three hypothesized processes is required for learning. Instead early map acquisition can occur with free exploration without target error correction, is based on sensory-to-sensory correspondences, and possible even for discontinuous maps. The findings are consistent with the idea that early sensorimotor map formation can involve instance-specific learning. NEW & NOTEWORTHY This study tested learning of novel sensorimotor maps in a variety of unusual circumstances, including learning a mapping that was permuted in such as way that it fragmented the sensorimotor workspace into discontinuous parts, thus not preserving sensory and motor topology. Participants could learn this mapping, and they could learn without motor outflow or targets. These results point to a robust learning mechanism building on individual instances, inspired from machine learning literature.

Feedback-dependent generalization

Generalization provides a window into the representational changes that occur during motor learning. Neural network models have been integral in revealing how the neural representation constrains the extent of generalization. Specifically, two key features are thought to define the pattern of generalization. First, generalization is constrained by the properties of the underlying neural units; with directionally tuned units, the extent of generalization is limited by the width of the tuning functions. Second, error signals are used to update a sensorimotor map to align the desired and actual output, with a gradient-descent learning rule ensuring that the error produces changes in those units responsible for the error. In prior studies, task-specific effects in generalization have been attributed to differences in neural tuning functions. Here we ask whether differences in generalization functions may arise from task-specific error signals. We systematically varied visual error information in a visuomotor adaptation task and found that this manipulation led to qualitative differences in generalization. A neural network model suggests that these differences are the result of error feedback processing operating on a homogeneous and invariant set of tuning functions. Consistent with novel predictions derived from the model, increasing the number of training directions led to specific distortions of the generalization function. Taken together, the behavioral and modeling results offer a parsimonious account of generalization that is based on the utilization of feedback information to update a sensorimotor map with stable tuning functions.