Continuous Gesture Control of a Robot Arm: Performance Is Robust to a Variety of Hand-to-Robot Maps.

OBJECTIVE: Despite advances in human-machine-interface design, we lack the ability to give people precise and fast control over high degree of freedom (DOF) systems, like robotic limbs. Attempts to improve control often focus on the static map that links user input to device commands; hypothesizing that the user's skill acquisition can be improved by finding an intuitive map. Here we investigate what map features affect skill acquisition. METHODS: Each of our 36 participants used one of three maps that translated their 19-dimensional finger movement into the 5 robot joints and used the robot to pick up and move objects. The maps were each constructed to maximize a different control principle to reveal what features are most critical for user performance. 1) Principal Components Analysis to maximize the linear capture of finger variance, 2) our novel Egalitarian Principal Components Analysis to maximize the equality of variance captured by each component and 3) a Nonlinear Autoencoder to achieve both high variance capture and less biased variance allocation across latent dimensions Results: Despite large differences in the mapping structures there were no significant differences in group performance. CONCLUSION: Participants' natural aptitude had a far greater effect on performance than the map. SIGNIFICANCE: Robot-user interfaces are becoming increasingly common and require new designs to make them easier to operate. Here we show that optimizing the map may not be the appropriate target to improve operator skill. Therefore, further efforts should focus on other aspects of the robot-user-interface such as feedback or learning environment.

Reflex voiding in rat occurs at consistent bladder volume regardless of pressure or infusion rate.

AIMS: The central nervous system (CNS) regulates lower urinary tract reflexes using information from sensory afferents; however, the mechanisms of this process are not well known. Pressure and volume were measured at the onset of the guarding and micturition reflexes across a range of infusion rates to provide insight into what the CNS is gauging to activate reflexes. METHODS: Female Sprague Dawley rats were anesthetized with urethane for open outlet cystometry. A set of 10 infusion rates (ranging 0.92-65.5 mL/h) were pseudo-randomly distributed across 30 single-fill cystometrograms. Bladder pressure and external urethral sphincter electromyography were used for the determination of the onset of the micturition and guarding reflexes, respectively. The bladder volume at the onset of both reflexes was estimated from the total infusion rate during a single fill. RESULTS: In response to many single-fill cystometrograms, there was an increased volume the bladder could store without a significant increase in pressure. Volume was adjusted for this effect for the analysis of how pressure and volume varied with infusion rate at the onset of the micturition and guarding reflexes. In 25 rats, the micturition reflex was evoked at similar volumes across all infusion rates, whereas the pressure at micturition reflex onset increased with increasing infusion rates. In 11 rats, the guarding reflex was evoked at similar pressures across infusion rates, but the volume decreased with increasing infusion rates. CONCLUSIONS: These results suggest that the CNS is interpreting volume from the bladder to activate the micturition reflex and pressure from the bladder to activate the guarding reflex.

Validation of a non-invasive, real-time, human-in-the-loop model of intracortical brain-computer interfaces.

Objective. Despite the tremendous promise of invasive brain-computer interfaces (iBCIs), the associated study costs, risks, and ethical considerations limit the opportunity to develop and test the algorithms that decode neural activity into a user's intentions. Our goal was to address this challenge by designing an iBCI model capable of testing many human subjects in closed-loop.Approach. We developed an iBCI model that uses artificial neural networks (ANNs) to translate human finger movements into realistic motor cortex firing patterns, which can then be decoded in real time. We call the model the joint angle BCI, or jaBCI. jaBCI allows readily recruited, healthy subjects to perform closed-loop iBCI tasks using any neural decoder, preserving subjects' control-relevant short-latency error correction and learning dynamics.Main results. We validated jaBCI offline through emulated neuron firing statistics, confirming that emulated neural signals have firing rates, low-dimensional PCA geometry, and rotational jPCA dynamics that are quite similar to the actual neurons (recorded in monkey M1) on which we trained the ANN. We also tested jaBCI in closed-loop experiments, our single study examining roughly as many subjects as have been tested world-wide with iBCIs (n= 25). Performance was consistent with that of the paralyzed, human iBCI users with implanted intracortical electrodes. jaBCI allowed us to imitate the experimental protocols (e.g. the same velocity Kalman filter decoder and center-out task) and compute the same seven behavioral measures used in three critical studies.Significance. These encouraging results suggest the jaBCI's real-time firing rate emulation is a useful means to provide statistically robust sample sizes for rapid prototyping and optimization of decoding algorithms, the study of bi-directional learning in iBCIs, and improving iBCI control.

S2 Heart Sound Detects Aortic Valve Calcification Independent of Hemodynamic Changes in Mice.

Background: Calcific aortic valve disease (CAVD) is often undiagnosed in asymptomatic patients, especially in underserved populations. Although artificial intelligence has improved murmur detection in auscultation exams, murmur manifestation depends on hemodynamic factors that can be independent of aortic valve (AoV) calcium load and function. The aim of this study was to determine if the presence of AoV calcification directly influences the S2 heart sound. Methods: Adult C57BL/6J mice were assigned to the following 12-week-long diets: (1) Control group (n = 11) fed a normal chow, (2) Adenine group (n = 4) fed an adenine-supplemented diet to induce chronic kidney disease (CKD), and (3) Adenine + HP (n = 9) group fed the CKD diet for 6 weeks, then supplemented with high phosphate (HP) for another 6 weeks to induce AoV calcification. Phonocardiograms, echocardiogram-based valvular function, and AoV calcification were assessed at endpoint. Results: Mice on the Adenine + HP diet had detectable AoV calcification (9.28 ± 0.74% by volume). After segmentation and dimensionality reduction, S2 sounds were labeled based on the presence of disease: Healthy, CKD, or CKD + CAVD. The dataset (2,516 S2 sounds) was split subject-wise, and an ensemble learning-based algorithm was developed to classify S2 sound features. For external validation, the areas under the receiver operating characteristic curve of the algorithm to classify mice were 0.9940 for Healthy, 0.9717 for CKD, and 0.9593 for CKD + CAVD. The algorithm had a low misclassification performance of testing set S2 sounds (1.27% false positive, 1.99% false negative). Conclusion: Our ensemble learning-based algorithm demonstrated the feasibility of using the S2 sound to detect the presence of AoV calcification. The S2 sound can be used as a marker to identify AoV calcification independent of hemodynamic changes observed in echocardiography.

Mathematical modeling of the lower urinary tract: A review.

AIMS: Understand what progress has been made toward a functionally predictive lower urinary tract (LUT) model, identify knowledge gaps, and develop from them a path forward. METHODS: We surveyed prominent mathematical models of the basic LUT components (bladder, urethra, and their neural control) and categorized the common modeling strategies and theoretical assumptions associated with each component. Given that LUT function emerges from the interaction of these components, we emphasized attempts to model their connections, and highlighted unmodeled aspects of LUT function. RESULTS: There is currently no satisfactory model of the LUT in its entirety that can predict its function in response to disease, treatment, or other perturbations. In particular, there is a lack of physiologically based mathematical descriptions of the neural control of the LUT. CONCLUSIONS: Based on our survey of the work to date, a potential path to a predictive LUT model is a modular effort in which models are initially built of individual tissue-level components using methods that are extensible and interoperable, allowing them to be connected and tested in a common framework. A modular approach will allow the larger goal of a comprehensive LUT model to be in sight while keeping individual efforts manageable, ensure new models can straightforwardly build on prior research, respect potential interactions between components, and incentivize efforts to model absent components. Using a modular framework and developing models based on physiological principles, to create a functionally predictive model is a challenge that the field is ready to undertake.

Cubature Kalman Filter Based Training of Hybrid Differential Equation Recurrent Neural Network Physiological Dynamic Models.

Modeling biological dynamical systems is challenging due to the interdependence of different system components, some of which are not fully understood. To fill existing gaps in our ability to mechanistically model physiological systems, we propose to combine neural networks with physics-based models. Specifically, we demonstrate how we can approximate missing ordinary differential equations (ODEs) coupled with known ODEs using Bayesian filtering techniques to train the model parameters and simultaneously estimate dynamic state variables. As a study case we leverage a well-understood model for blood circulation in the human retina and replace one of its core ODEs with a neural network approximation, representing the case where we have incomplete knowledge of the physiological state dynamics. Results demonstrate that state dynamics corresponding to the missing ODEs can be approximated well using a neural network trained using a recursive Bayesian filtering approach in a fashion coupled with the known state dynamic differential equations. This demonstrates that dynamics and impact of missing state variables can be captured through joint state estimation and model parameter estimation within a recursive Bayesian state estimation (RBSE) framework. Results also indicate that this RBSE approach to training the NN parameters yields better outcomes (measurement/state estimation accuracy) than training the neural network with backpropagation through time in the same setting.

Age is associated with reduced urethral pressure and afferent activity in rat.

Age-related changes in the lower urinary tract (LUT) can affect the coordination of reflexes and increase the incidence of bladder disorders in elderly. This study examines the age-related loss of urethral signaling capability by measuring the afferent activity directly. We find that less urethral pressure develops in response to fluid flow in old rats compared to young rats and that pressure and flow evoke less urethral afferent activation. These findings are consistent with our previous study demonstrating that the urethra-to-bladder reflex, which is required for efficient voiding, becomes weaker with age. We measured the pudendal afferent response in young (4-7 months) and old (18-24 months) rats to fluid flow in the urethra across a range of flow rates. We used paraffin embedding and hematoxylin and eosin staining to quantify age-related changes in the sensory branch of the pudendal nerve. Urethral afferent signaling in response to the same urethral flow rates was weaker in older animals. That is, the sensitivity of urethra afferents to flow decreased with age, and higher flow rates were required in older animals to recruit urethra afferents. There was also a reduction in the myelin thickness of pudendal afferents in old rats, which is a possible contributing factor to the sensory activity. Furthermore, the same flow rates evoked less pressure in the urethras of old animals, indicating there is an age-related change of the urethral tissue that reduces the pressure stimulus to which these afferents respond. These results help characterize the underlying changes in LUT system with age.

Sensitivity of urethral flow-evoked voiding reflexes decline with age in the rat: insights into age-related underactive bladder.

The prevalence of underactive bladder (UAB) increases with age, suggesting a link between age-related processes and lower urinary tract (LUT) symptoms; however, the underlying mechanisms of age-related UAB are poorly understood. Understanding how aging affects LUT reflexes may help in the development of new treatments by identifying mechanistic targets. In this work, we studied the relationship between age and systems-level function of the LUT and tested the hypothesis that aging is related to weakening of reflexes that control voiding. Three groups of anesthetized female rats, young (4-7 mo old), mature (11-14 mo old), and old (18-24 mo old), were used to quantify the effect of aging on LUT reflexes. A double-lumen catheter enabled us to control the bladder volume and urethral flow rate independently, under quasi-isovolumetric bladder conditions. We systematically investigated the reflex bladder contractions evoked by combinations of rates of urethral infusion and bladder fill volumes as a function of age. Urethral infusion with the same flow rate evoked bladder contractions (via the augmenting reflex) in old animals less often than in younger animals. Furthermore, old animals needed more fluid in their bladders (relative to their bladder capacity) before urethra flow-evoked bladder contractions could be triggered at all, suggesting a delay in the switch of the LUT to "voiding mode." Old rats also showed longer and weaker bladder contractions than young or mature rats. Taken together, this suggests there is an age-related functional weakening and loss of sensitivity in LUT reflexes, which may contribute to age-related UAB symptoms.

Multiple Reflex Pathways Contribute to Bladder Activation by Intraurethral Stimulation in Persons With Spinal Cord Injury.

OBJECTIVE: To measure the urodynamic effects of electrical co-stimulation of 2 individual sites in the proximal and distal urethra in persons with spinal cord injury (SCI). This work was motivated by preclinical findings that selective co-stimulation of the cranial urethral sensory nerve and the dorsal genital nerve, which innervate the proximal and distal portions of the urethra, respectively, increased reflex bladder activation and voiding efficiency. MATERIALS AND METHODS: Electrical co-stimulation of urethral afferents was conducted in persons with chronic SCI during urodynamics. The effects of different frequencies of intraurethral stimulation at multiple urethral locations on bladder pressure and pelvic floor electromyographic activity were measured. RESULTS: Electromyographic activity indicated that multiple reflex pathways were recruited through stimulation that contributed to bladder activation. The size of reflex bladder contractions evoked by stimulation was dependent on stimulation location or reflex activated and stimulation frequency. CONCLUSION: Pudendal nerve afferents are a promising target to restore lost bladder control, as stimulation with different frequencies may be used to treat urinary incontinence and increase continent volumes or to generate stimulation-evoked bladder contractions for on-demand voiding. This work identified that co-stimulation of multiple afferent reflex pathways can enhance activation of spinal circuits and may enable improved bladder emptying in SCI when stimulation of a single pathway is not sufficient.

Sensory feedback from the urethra evokes state-dependent lower urinary tract reflexes in rat.

KEY POINTS: The lower urinary tract is regulated by reflexes responsible for maintaining continence and producing efficient voiding. It is unclear how sensory information from the bladder and urethra engages differential, state-dependent reflexes to either maintain continence or promote voiding. Using a new in vivo experimental approach, we quantified how sensory information from the bladder and urethra are integrated to switch reflex responses to urethral sensory feedback from maintaining continence to producing voiding. The results demonstrate how sensory information regulates state-dependent reflexes in the lower urinary tract and contribute to our understanding of the pathophysiology of urinary retention and incontinence where sensory feedback may engage these reflexes inappropriately. ABSTRACT: Lower urinary tract reflexes are mediated by peripheral afferents from the bladder (primarily in the pelvic nerve) and the urethra (in the pudendal and pelvic nerves) to maintain continence or initiate micturition. If fluid enters the urethra at low bladder volumes, reflexes relax the bladder and evoke external urethral sphincter (EUS) contraction (guarding reflex) to maintain continence. Conversely, urethral flow at high bladder volumes, excites the bladder (micturition reflex) and relaxes the EUS (augmenting reflex). We conducted measurements in a urethane-anaesthetized in vivo rat preparation to characterize systematically the reflexes evoked by fluid flow through the urethra. We used a novel preparation to manipulate sensory feedback from the bladder and urethra independently by controlling bladder volume and urethral flow. We found a distinct bladder volume threshold (74% of bladder capacity) above which flow-evoked bladder contractions were 252% larger and evoked phasic EUS activation 2.6 times as often as responses below threshold, clearly demonstrating a discrete transition between continence (guarding) and micturition (augmenting) reflexes. Below this threshold urethral flow evoked tonic EUS activity, indicative of the guarding reflex, that was proportional to the urethral flow rate. These results demonstrate the complementary roles of sensory feedback from the bladder and urethra in regulating reflexes in the lower urinary tract that depend on the state of the bladder. Understanding the neural control of functional reflexes and how they are mediated by sensory information in the bladder and urethra will open new opportunities, especially in neuromodulation, to treat pathologies of the lower urinary tract.

Loss of Ranbp2 in motoneurons causes disruption of nucleocytoplasmic and chemokine signaling, proteostasis of hnRNPH3 and Mmp28, and development of amyotrophic lateral sclerosis-like syndromes.

The pathogenic drivers of sporadic and familial motor neuron disease (MND), such amyotrophic lateral sclerosis (ALS), are unknown. MND impairs the Ran GTPase cycle, which controls nucleocytoplasmic transport, ribostasis and proteostasis; however, cause-effect mechanisms of Ran GTPase modulators in motoneuron pathobiology have remained elusive. The cytosolic and peripheral nucleoporin Ranbp2 is a crucial regulator of the Ran GTPase cycle and of the proteostasis of neurological disease-prone substrates, but the roles of Ranbp2 in motoneuron biology and disease remain unknown. This study shows that conditional ablation of Ranbp2 in mouse Thy1 motoneurons causes ALS syndromes with hypoactivity followed by hindlimb paralysis, respiratory distress and, ultimately, death. These phenotypes are accompanied by: a decline in the nerve conduction velocity, free fatty acids and phophatidylcholine of the sciatic nerve; a reduction in the g-ratios of sciatic and phrenic nerves; and hypertrophy of motoneurons. Furthermore, Ranbp2 loss disrupts the nucleocytoplasmic partitioning of the import and export nuclear receptors importin ß and exportin 1, respectively, Ran GTPase and histone deacetylase 4. Whole-transcriptome, proteomic and cellular analyses uncovered that the chemokine receptor Cxcr4, its antagonizing ligands Cxcl12 and Cxcl14, and effector, latent and activated Stat3 all undergo early autocrine and proteostatic deregulation, and intracellular sequestration and aggregation as a result of Ranbp2 loss in motoneurons. These effects were accompanied by paracrine and autocrine neuroglial deregulation of hnRNPH3 proteostasis in sciatic nerve and motoneurons, respectively, and post-transcriptional downregulation of metalloproteinase 28 in the sciatic nerve. Mechanistically, our results demonstrate that Ranbp2 controls nucleocytoplasmic, chemokine and metalloproteinase 28 signaling, and proteostasis of substrates that are crucial to motoneuronal homeostasis and whose impairments by loss of Ranbp2 drive ALS-like syndromes.

Estimating postvoid residual volume without measuring residual bladder volume during serial cystometrograms.

The postvoid residual volume (PVR) is a common urodynamic parameter used to quantify the severity of lower urinary tract dysfunction. However, the serial cystometrograms that are typically used to assess bladder function in animal models make measuring PVR very difficult. Current approaches are to either remove PVR after each void to measure it, which is disruptive to the bladder, or to neglect the unknown contribution to PVR from ureter flow, which results in inaccurate estimates. We propose a procedure to estimate PVR during a serial cystometrogram that requires only a single measurement, rather than measuring after each void. Moreover, this measurement can occur at the end of the experiment such that it does not affect the bladder during data collection. We mathematically express PVR for all voids during a serial cystometrogram using a linear recurrence equation and use this equation to build an estimation procedure for PVR. Using in vivo measurements in urethane anesthetized rats and computer simulations we show that the estimation procedure is at least as accurate in determining PVR as the traditional method of measuring PVR after each void. Furthermore, we demonstrate the adverse effects of repeated PVR measurements in a common animal model of cystitis. Using the proposed procedure can increase the efficiency and accuracy of determining PVR for a serial cystometrogram and is less disruptive to the system under study. This, in turn, allows the calculation of other urodynamic parameters that are critical for research studies, including voiding efficiency and bladder capacity.

Sensory and circuit mechanisms mediating lower urinary tract reflexes.

Neural control of continence and micturition is distributed over a network of interconnected reflexes. These reflexes integrate sensory information from the bladder and urethra and are modulated by descending influences to produce different physiological outcomes based on the information arriving from peripheral afferents. Therefore, the mode of activation of primary afferents is essential in understanding the action of spinal reflex pathways in the lower urinary tract. We present an overview of sensory mechanisms in the bladder and urethra focusing on their spinal integration, identify the cardinal spinal reflexes responsible for continence and micturition, and describe how their functional role is controlled via peripheral afferent activity.

Dynamics of the sensory response to urethral flow over multiple time scales in rat.

KEY POINTS: Sensory information from the urethra is essential to maintain continence and to achieve efficient micturition and when compromised by disease or injury can lead to substantial loss of function. Despite the key role urethral sensory information plays in the lower urinary tract, the relationship between physiological urethral stimuli, such as fluid flow, and the neural sensory response is poorly understood. This work systematically quantifies pudendal afferent responses to a range of fluid flows in the urethra in vivo and describes a previously unknown long-term neural accommodation phenomenon in these afferents. We present a compact mechanistic mathematical model that reproduces the pudendal sensory activity in response to urethral flow. These results have implications for understanding urinary tract dysfunction caused by neuropathy or nerve damage, such as urinary retention or incontinence, as well as for the development of strategies to mitigate the symptoms of these conditions. The pudendal nerve carries sensory information from the urethra that controls spinal reflexes necessary to maintain continence and achieve efficient micturition. Despite the key role urethral sensory feedback plays in regulation of the lower urinary tract, there is little information about the characteristics of urethral sensory responses to physiological stimuli, and the quantitative relationship between physiological stimuli and the evoked sensory activation is unknown. Such a relation is critical to understanding the neural control of the lower urinary tract and how dysfunction arises in disease states. We systematically quantified pudendal afferent responses to fluid flow in the urethra in vivo in the rat. We characterized the sensory response across a range of stimuli, and describe a previously unreported long-term neural accommodation phenomenon. We developed and validated a compact mechanistic mathematical model capable of reproducing the pudendal sensory activity in response to arbitrary profiles of urethral flows. These results describe the properties and function of urethral afferents that are necessary to understand how sensory disruption manifests in lower urinary tract pathophysiology.

A neuron model of stochastic resonance using rectangular pulse trains.

Stochastic resonance (SR) is the enhanced representation of a weak input signal by the addition of an optimal level of broadband noise to a nonlinear (threshold) system. Since its discovery in the 1980s the domain of input signals shown to be applicable to SR has greatly expanded, from strictly periodic inputs to now nearly any aperiodic forcing function. The perturbations (noise) used to generate SR have also expanded, from white noise to now colored noise or vibrational forcing. This study demonstrates that a new class of perturbations can achieve SR, namely, series of stochastically generated biphasic pulse trains. Using these pulse trains as 'noise' we show that a Hodgkin Huxley model neuron exhibits SR behavior when detecting weak input signals. This result is of particular interest to neuroscience because nearly all artificial neural stimulation is implemented with square current or voltage pulses rather than broadband noise, and this new method may facilitate the translation of the performance gains achievable through SR to neural prosthetics.

A spinal GABAergic mechanism is necessary for bladder inhibition by pudendal afferent stimulation.

Electrical stimulation of pudendal afferents can inhibit bladder contractions and increase bladder capacity. Recent results suggest that stimulation-evoked bladder inhibition is mediated by a mechanism other than activation of sympathetic bladder efferents in the hypogastric nerve, generating α-adrenergic receptor-mediated inhibition at the vesical ganglia and/or ß-adrenergic receptor-mediated direct inhibition of the detrusor muscle. We investigated several inhibitory neurotransmitters that may instead be necessary for stimulation-evoked inhibition and found that intravenous picrotoxin, a noncompetitive GABAA antagonist, significantly and reversibly blocked pudendal afferent stimulation-evoked inhibition of bladder contractions in a dose-dependent manner. Similarly, intravenous picrotoxin also blocked pudendal afferent stimulation-evoked inhibition of nociceptive bladder contractions evoked by acetic acid infusion. Furthermore, intrathecal administration of picrotoxin at the lumbosacral spinal cord also blocked bladder inhibition by pudendal afferent stimulation. On the other hand, glycinergic, adrenergic, or opioidergic mechanisms were not necessary for bladder inhibition evoked by pudendal afferent stimulation. These results identify a lumbosacral spinal GABAergic mechanism of bladder inhibition evoked by pudendal afferent stimulation.

A reductionist approach to the analysis of learning in brain-computer interfaces.

The complexity and scale of brain-computer interface (BCI) studies limit our ability to investigate how humans learn to use BCI systems. It also limits our capacity to develop adaptive algorithms needed to assist users with their control. Adaptive algorithm development is forced offline and typically uses static data sets. But this is a poor substitute for the online, dynamic environment where algorithms are ultimately deployed and interact with an adapting user. This work evaluates a paradigm that simulates the control problem faced by human subjects when controlling a BCI, but which avoids the many complications associated with full-scale BCI studies. Biological learners can be studied in a reductionist way as they solve BCI-like control problems, and machine learning algorithms can be developed and tested in closed loop with the subjects before being translated to full BCIs. The method is to map 19 joint angles of the hand (representing neural signals) to the position of a 2D cursor which must be piloted to displayed targets (a typical BCI task). An investigation is presented on how closely the joint angle method emulates BCI systems; a novel learning algorithm is evaluated, and a performance difference between genders is discussed.

The influence of visual motion on motor learning.

How does visual perception shape the way we coordinate movements? Recent studies suggest that the brain organizes movements based on minimizing reaching errors in the presence of motor and sensory noise. We present an alternative hypothesis in which movement trajectories also result from acquired knowledge about the geometrical properties of the object that the brain is controlling. To test this hypothesis, we asked human subjects to control a simulated kinematic linkage by continuous finger motion, a completely novel experience. This paradigm removed all biases arising from influences of limb dynamics and past experience. Subjects were exposed to two different types of visual feedback; some saw the entire simulated linkage and others saw only the moving extremity. Consistent with our hypothesis, subjects learned to move the simulated linkage along geodesic lines corresponding to the geometrical structure of the observed motion. Thus, optimizing final accuracy is not the unique determinant of trajectory formation.

Sensory motor remapping of space in human-machine interfaces.

Studies of adaptation to patterns of deterministic forces have revealed the ability of the motor control system to form and use predictive representations of the environment. These studies have also pointed out that adaptation to novel dynamics is aimed at preserving the trajectories of a controlled endpoint, either the hand of a subject or a transported object. We review some of these experiments and present more recent studies aimed at understanding how the motor system forms representations of the physical space in which actions take place. An extensive line of investigations in visual information processing has dealt with the issue of how the Euclidean properties of space are recovered from visual signals that do not appear to possess these properties. The same question is addressed here in the context of motor behavior and motor learning by observing how people remap hand gestures and body motions that control the state of an external device. We present some theoretical considerations and experimental evidence about the ability of the nervous system to create novel patterns of coordination that are consistent with the representation of extrapersonal space. We also discuss the perspective of endowing human-machine interfaces with learning algorithms that, combined with human learning, may facilitate the control of powered wheelchairs and other assistive devices.