External Sensory Feedback Research Articles

Movement predictability modulates sensorimotor processing.

An important factor for optimal sensorimotor control is how well we are able to predict sensory feedback from internal and external sources during movement. If predictability decreases due to external disturbances, the brain is able to adjust muscle activation and the filtering of incoming sensory inputs. However, little is known about sensorimotor adjustments when predictability is increased by availability of additional internal feedback. In the present study we investigated how modifications of internal and external sensory feedback influence the control of muscle activation and gating of sensory input. Co-activation of forearm muscles, somatosensory evoked potentials (SEP) and short afferent inhibition (SAI) were assessed during three object manipulation tasks designed to differ in the predictability of sensory feedback. These included manipulation of a shared object with both hands (predictable coupling), manipulation of two independent objects without (uncoupled) and with external interference on one of the objects (unpredictable coupling). We found a task-specific reduction in co-activation during the predictable coupling compared to the other tasks. Less sensory gating, reflected in larger subcortical SEP amplitudes, was observed in the unpredictable coupling task. SAI behavior was closely linked to the subcortical SEP component indicating an important function of subcortical sites in predictability related SEP gating and their direct influence on M1 inhibition. Together, these findings suggest that the unpredictable coupling task cannot only rely on predictive forward control and is compensated by enhancing co-activation and increasing the saliency for external stimuli by reducing sensory gating at subcortical level. This behavior might serve as a preparatory step to compensate for external disturbances and to enhance processing and integration of all incoming external stimuli to update the current sensorimotor state. In contrast, predictive forward control is accurate in the predictable coupling task due to the integrated sensory feedback from both hands where sensorimotor resources are economized by reducing muscular co-activation and increasing sensory gating.

The Role of the Cerebellar and Vestibular Networks in Anxiety Disorders and Depression: the Internal Model Hypothesis

Clinical data and animal studies confirmed that the cerebellum and the vestibular system are involved in emotions. Nowadays, no real consensus has really emerged to explain the clinical symptoms in humans and behavioral deficits in the animal models. We envisage here that the cerebellum and the vestibular system play complementary roles in emotional reactivity. The cerebellum integrates a large variety of exteroceptive and proprioceptive information necessary to elaborate and to update the internal model: in emotion, as in motor processes, it helps our body and self to adapt to the environment, and to anticipate any changes in such environment in order to produce a time-adapted response. The vestibular system provides relevant environmental stimuli (i.e., gravity, self-position, and movement) and is involved in self-perception. Consequently, cerebellar or vestibular disorders could generate « internal fake news» (due to lack or false sensory information and/or integration) that could, in turn, generate potential internal model deficiencies. In this case, the alterations provoke false anticipation of motor command and external sensory feedback, associated with unsuited behaviors. As a result, the individual becomes progressively unable to cope with the environmental solicitation. We postulate that chronically unsuited, and potentially inefficient, behavioral and visceral responses to environmental solicitations lead to stressful situations. Furthermore, this inability to adapt to the context of the situation generates chronic anxiety which could precede depressive states.

Continuous Identification of Freezing of Gait in Parkinson’s Patients Using Artificial Neural Networks and Instrumented Shoes

Freezing of gait is an episodic phenomena faced by many patients with Parkinson's disease. It is characterized by episodes during which patients are unable to generate effective forward stepping movements, despite absence of motor deficits. During the onset of the event, the patients are less stable with statistically different stride width, toe in/out angle and center of pressure distance. It has been postulated that the degree of freezing can be reduced by providing external sensory feedback to the patients during the event. However, this intervention could be facilitated by accurate identification of freezing events in real-time. This manuscript presents an Artificial Neural Network model which uses signals recorded by an instrumented footwear to predict if a walking subject is having a freezing episode. Our model presented in this paper is capable of continuously predicting freezing of gait events at a high temporal resolution of 50 Hz, using a 0.5 second window of data recorded by the instrumented shoes, with a sensitivity of 96.0±2.5%, a specificity of 99.6±0.3%, a precision of 89.5±5.9%, and an accuracy of 99.5±0.4%. This algorithm was tested with data collected from 10 patients with Parkinson's disease with frequent freezing of gait episodes.

Consolidation and retention of motor skill after motor imagery training

Complex motor tasks are learned through training which results in lasting improvement in sensorimotor performance and accuracy. Learning a motor skill is commonly attained via physical execution. However, research has shown that cognitive training, such as motor imagery (MI), effectively facilitates skill learning. Neurophysiological findings suggest that learning-induced plasticity in the human motor cortex, subserving consolidation and retention of motor skills, is stronger after movement execution (ME) than after MI training. Here, we designed an experimental task able to test for the fast and slow learning phases and for retention of motor skills for both MI and ME. We hypothesize that differences between MI and ME training would emerge in terms of reduced consolidation and retention of motor skills. Twenty-four young healthy subjects were divided into two groups, performing MI or ME training. Participants wore sensor-engineered gloves and their sensorimotor performance was assessed over a period of 15 days with 4-days training. We analysed the touch duration (TD), the inter-tapping interval (ITI), movement rate and accuracy. Results showed that (i) during the first phase of acquisition of motor skills, sensorimotor performance improved similarly in MI and ME groups; (ii) during the second learning phase movement rate increased more in ME than MI group and this difference was mainly driven by differences in the duration of TD; (iii) consolidation deficits with MI training reflected in impaired retention of the acquired skills, as TD and ITI were larger and movement rate was lower in the MI group with respect to the ME, till to 10 days after the last training session. Explicit component of motor learning, accuracy, was maintained in retention phase in both groups. Following our hypothesis, our findings show that MI training is as effective as ME within the first learning phase, but consolidation and retention of motor skills are less effective following MI training. This study highlights MI limitations and suggests option to enhance MI, as by providing an external sensory feedback.

Effects can precede their cause in the sense of agency

The sense of agency, i.e., the feeling that one׳s action is the cause of an external sensory event, involves causal inference based on the predicted sensory outcome of a motor act. Here, we investigated whether this inference process faithfully implements the physical principle that a cause (motor act) temporally precedes its effect (external sensory feedback). To this end, we presented participants with visual flashes that were temporally offset from voluntary button presses, including scenarios where the flash occurred shortly before the press. Participants then judged their experience of agency. As expected, cause-effect order is an important cue for this task: participants were far more likely to report agency for temporally lagging flashes than for leading flashes, even if very long sensory delays also disrupted the sense of agency (Experiment 1). This suggests that the temporal order between action and sensation is the dominant temporal cue for agency. However, when participants judged whether they had caused a first flash that occurred before the button press or a second flash that occurred afterwards, the temporal threshold for rejecting leading first flashes was relaxed proportionally to the delay of the second flash (Experiment 2). There was competition between different sensorimotor timing cues (temporal order favored the second flash and temporal proximity favored the first flash), and participants׳ tolerance for cause-effect inversions was modulated by the strength of the later, conflicting cue. We conclude that the perceived order of action and sensation is not used in a winner-take-all fashion in inference of agency. Instead, a probabilistic negotiation of the different timing cues favoring different flash events takes place postdictively, after presentation of the second flash.

Discretization and fitting of nominal data for autonomous robots

This paper presents methodologies to discretize nominal robot paths extracted from 3-D CAD drawings. Behind robot path discretization is the ability to have a robot adjusting the traversed paths so that the contact between robot tool and work-piece is properly maintained. In addition, a hybrid force/motion control system based on Fuzzy-PI control is proposed to adjust robot paths with external sensory feedback. All these capabilities allow to facilitate the robot programming process and to increase the robot’s autonomy.

Direct Digital Robot Control using a Force-Torque Sensor

The paper treats the analysis and design of external sensory feedback loops in robots. Main emphasis is laid on force-torque sensors but the techniques described are more general. It is shown that by using very simplified dynamical robot models the main effects occuring in a robot’s reaction with its environment are amenable to a systematic treatment. Timing schemes based on the interactions between the robot’s internal microprocessor and a supervisory control microprocessor are outlined and related to sampled data theory. Digital control techniques using z-transform and state space methods are applied to typical problems as are “moving to and pressing onto a surface with prescribed force”, “motion speed control in a grinding process”, and “robot teaching via force-torque-sensors”.