A real-time, practical sensor fault-tolerant module for robust EMG pattern recognition

Swing-phase detection of locomotive mode transitions for smooth multi-functional robotic lower-limb prosthesis control

Robotic lower-limb prostheses, with their actively powered joints, may significantly improve amputee users' mobility and enable them to obtain healthy-like gait in various modes of locomotion in daily life. However, timely recognition of the amputee users' locomotive mode and mode transition still remains a major challenge in robotic lower-limb prosthesis control. In the paper, the authors present a new multi-dimensional dynamic time warping (mDTW)-based intent recognizer to provide high-accuracy recognition of the locomotion mode/mode transition sufficiently early in the swing phase, such that the prosthesis' joint-level motion controller can operate in the correct locomotive mode and assist the user to complete the desired (and often power-demanding) motion in the stance phase. To support the intent recognizer development, the authors conducted a multi-modal gait data collection study to obtain the related sensor signal data in various modes of locomotion. The collected data were then segmented into individual cycles, generating the templates used in the mDTW classifier. Considering the large number of sensor signals available, we conducted feature selection to identify the most useful sensor signals as the input to the mDTW classifier. We also augmented the standard mDTW algorithm with a voting mechanism to make full use of the data generated from the multiple subjects. To validate the proposed intent recognizer, we characterized its performance using the data cumulated at different percentages of progression into the gait cycle (starting from the beginning of the swing phase). It was shown that the mDTW classifier was able to recognize three locomotive mode/mode transitions (walking, walking to stair climbing, and walking to stair descending) with 99.08% accuracy at 30% progression into the gait cycle, well before the stance phase starts. With its high performance, low computational load, and easy personalization (through individual template generation), the proposed mDTW intent recognizer may become a highly useful building block of a prosthesis control system to facilitate the robotic prostheses' real-world use among lower-limb amputees.

Implications of EMG channel count: enhancing pattern recognition online prosthetic testing

Introduction

Myoelectric pattern recognition systems have shown promising control of upper limb powered prostheses and are now commercially available. These pattern recognition systems typically record from up to 8 muscle sites, whereas other control systems use two-site control. While previous offline studies have shown 8 or fewer sites to be optimal, real-time control was not evaluated.

Methods

Six individuals with no limb absence and four individuals with a transradial amputation controlled a virtual upper limb prosthesis using pattern recognition control with 8 and 16 channels of EMG. Additionally, two of the individuals with a transradial amputation performed the Assessment for Capacity of Myoelectric Control (ACMC) with a multi-articulating hand and wrist prosthesis with the same channel count conditions.

Results

Users had significant improvements in control when using 16 compared to 8 EMG channels including decreased classification error (p = 0.006), decreased completion time (p = 0.019), and increased path efficiency (p = 0.013) when controlling a virtual prosthesis. ACMC scores increased by more than three times the minimal detectable change from the 8 to the 16-channel condition.

Discussion

The results of this study indicate that increasing EMG channel count beyond the clinical standard of 8 channels can benefit myoelectric pattern recognition users.

A Review of Current State-of-the-Art Control Methods for Lower-Limb Powered Prostheses

Lower-limb prostheses aim to restore ambulatory function for individuals with lower-limb amputations. While the design of lower-limb prostheses is important, this paper focuses on the complementary challenge - the control of lower-limb prostheses. Specifically, we focus on powered prostheses, a subset of lower-limb prostheses, which utilize actuators to inject mechanical power into the walking gait of a human user. In this paper, we present a review of existing control strategies for lower-limb powered prostheses, including the control objectives, sensing capabilities, and control methodologies. We separate the various control methods into three main tiers of prosthesis control: high-level control for task and gait phase estimation, mid-level control for desired torque computation (both with and without the use of reference trajectories), and low-level control for enforcing the computed torque commands on the prosthesis. In particular, we focus on the high- and mid-level control approaches in this review. Additionally, we outline existing methods for customizing the prosthetic behavior for individual human users. Finally, we conclude with a discussion on future research directions for powered lower-limb prostheses based on the potential of current control methods and open problems in the field.

Characterizing Prosthesis Control Fault during Human-Prosthesis Interactive Walking Using Intrinsic Sensors

The physical interactions between wearable lower limb robots and humans have been investigated to inform effective robot design for walking augmentation. However, human-robot interactions when internal faults occur within robots have not been systematically reported, but it is essential to improve the robustness of robotic devices and ensure the user's safety. This paper aims to (1) present a methodology to characterize the behavior of the robotic transfemoral prosthesis as an effective wearable robot platform while interacting with the users in the presence of internal faults, and (2) identify the potential data sources for accurate detection of the prosthesis fault. We first obtained the human perceived response in terms of their walking stability when the prosthesis control fault (inappropriate intrinsic control output/command) was emulated/applied in level-ground walking. Then the measurements and their features, obtained from the transfemoral prosthesis, were examined for the emulated faults that elicited a sense of instability in human users. The optimal features that contributed the most in separating faulty interaction from the normal walking condition were determined using two machine-learning-based approaches: One-Class Support Vector Machine (OCSVM) and Mahalanobis Distance (MD) classifier. The OCSVM anomaly detector could achieve an average sensitivity of 85.7 % and an average false alarm rate of 1.7 % with a reasonable detecting time of 147.6 ms for detecting emulated control errors among all subjects. The result demonstrates the potential of using machine-learning-based schemes in identifying prosthesis control faults based on intrinsic sensors on the prosthesis. This study presents a procedure to study human-robot fault tolerance and inform the future design of robust prosthesis control.

Intrinsically Stretchable Block Copolymer Based on PEDOT:PSS for Improved Performance in Bioelectronic Applications

The conductive polyelectrolyte complex poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) is ubiquitous in research dealing with organic electronic devices (e.g., solar cells, wearable and implantable sensors, and electrochemical transistors). In many bioelectronic applications, the applicability of commercially available formulations of PEDOT:PSS (e.g., Clevios) is limited by its poor mechanical properties. Additives can be used to increase the compliance but pose a risk of leaching, which can result in device failure and increased toxicity (in biological settings). Thus, to increase the mechanical compliance of PEDOT:PSS without additives, we synthesized a library of intrinsically stretchable block copolymers. In particular, controlled radical polymerization using a reversible addition-fragmentation transfer process was used to generate block copolymers consisting of a block of PSS (of fixed length) appended to varying blocks of poly(poly(ethylene glycol) methyl ether acrylate) (PPEGMEA). These block copolymers (PSS₍₁₎-b-PPEGMEA_(x), where x ranges from 1 to 6) were used as scaffolds for oxidative polymerization of PEDOT. By increasing the lengths of the PPEGMEA segments on the PEDOT:[PSS₍₁₎-b-PPEGMEA_(1-6)] block copolymers, ("Block-1" to "Block-6"), or by blending these copolymers with PEDOT:PSS, the mechanical and electronic properties of the polymer can be tuned. Our results indicate that the polymer with the longest block of PPEGMEA, Block-6, had the highest fracture strain (75%) and lowest elastic modulus (9.7 MPa), though at the expense of conductivity (0.01 S cm^-1). However, blending Block-6 with PEDOT:PSS to compensate for the insulating nature of the PPEGMEA resulted in increased conductivity [2.14 S cm^-1 for Blend-6 (2:1)]. Finally, we showed that Block-6 outperforms a commercial formulation of PEDOT:PSS as a dry electrode for surface electromyography due to its favorable mechanical properties and better adhesion to skin.

Design and Validation of a Sensor Fault-Tolerant Module for Real-Time High-Density EMG Pattern Recognition

With the advancements in electronics technology, high-density (HD) EMG sensing systems have become available and have been investigated for their feasibility and performance in neural-machine interface (NMI) applications. Comparing to the traditional single channel-based targeted muscle sensing method, HD EMG sensing performs a sampling of the electrical activity over a larger surface area and has the promise of 1) providing richer neural information from one temporal and two spatial dimensions and 2) ease of wear in real life without the need of anatomically targeted electrode placement. To use HD EMG in real-time NMI applications, challenges including high computational burden and unreliability of EMG recordings over time need to be addressed. This paper presented an HD EMG PR based NMI which seamlessly integrates HD EMG PR with a Sensor Fault-Tolerant Module (SFTM) which aimed to provide robust PR in real time. Experimental results showed that the SFTM was able to recover the PR accuracies by 6%-22% from disturbances including contact artifacts and loose contacts. A Python-based implementation of the proposed HD EMG SFTM was developed and was demonstrated to be computationally efficient for real-time performance. These results have demonstrated the feasibility of a robust real-time HD EMG PR-based NMI.

Myoelectric control of robotic lower limb prostheses: a review of electromyography interfaces, control paradigms, challenges and future directions

Objective.Advanced robotic lower limb prostheses are mainly controlled autonomously. Although the existing control can assist cyclic movements during locomotion of amputee users, the function of these modern devices is still limited due to the lack of neuromuscular control (i.e. control based on human efferent neural signals from the central nervous system to peripheral muscles for movement production). Neuromuscular control signals can be recorded from muscles, called electromyographic (EMG) or myoelectric signals. In fact, using EMG signals for robotic lower limb prostheses control has been an emerging research topic in the field for the past decade to address novel prosthesis functionality and adaptability to different environments and task contexts. The objective of this paper is to review robotic lower limb Prosthesis control via EMG signals recorded from residual muscles in individuals with lower limb amputations.Approach.We performed a literature review on surgical techniques for enhanced EMG interfaces, EMG sensors, decoding algorithms, and control paradigms for robotic lower limb prostheses.Main results.This review highlights the promise of EMG control for enabling new functionalities in robotic lower limb prostheses, as well as the existing challenges, knowledge gaps, and opportunities on this research topic from human motor control and clinical practice perspectives.Significance.This review may guide the future collaborations among researchers in neuromechanics, neural engineering, assistive technologies, and amputee clinics in order to build and translate true bionic lower limbs to individuals with lower limb amputations for improved motor function.

User Independent Estimations of Gait Events With Minimal Sensor Data

GOAL

The purpose of this study was to provide an initial examination of the utility of the Beta Process - Auto Regressive - Hidden Markov Model (BP-AR-HMM) for the prior identification of gait events. A secondary objective was to determine whether the output of the model could be used for classification and prediction of locomotion states.

METHODS

In this study we utilized the output of the BP-AR-HMM to develop user-independent identification of gait events and gait classification from an idealized three-dimensional acceleration signal. The input acceleration data were collected from two walking (1.4 and 1.6 ms^-1) and two running (2.6 and 3.0 ms^-1) steady state speeds, and during two dynamic walk to run and run to walk transitions (1.8-2.4 and 2.4-1.8 ms^-1) on an instrumented force treadmill.

RESULTS

The BP-AR-HMM identified 9 unique states. Of these, two states, 4 and 1, were utilized to estimate initial contact and toe off, respectively. The lead time from the first instance of state 4 to initial contact was 0.13 ± 0.02 s. Similarly, the first instance of state 1 occurred 0.14 ± 0.03 s before toe off. Two other states (3 and 7) were examined for possible utilization in a probabilistic model for the prediction of pending locomotion state transitions.

CONCLUSION

The identification of gait events prior to their occurrence by the BP-AR-HMM appears to be an approach that can minimize the quantity of sensor data in an offline approach. Furthermore, there is evidence it could also be used as a basis to build a probabilistic model to estimate locomotion transitions.

Recurrent Neural Network for Contaminant Type Detector in Surface Electromyography Signals

A surface Electromyography (sEMG) contaminant type detector has been developed by using a Recurrent Neural Network (RNN) with Long Short-Term (LSMT) units in its hidden layer. This setup may reduce the contamination detection processing time since there is no need for feature extraction so that the classification occurs directly from the sEMG signal. The publicly available NINAPro (Non-Invasive Adaptive Prosthetics) database sEMG signals was used to train and test the network. Signals were contaminated with White Gaussian Noise, Movement Artifact, ECG and Power Line Interference. Two out of the 40 healthy subjects' data were considered to train the network and the other 38 to test it. Twelve models were trained under a -20dB contamination, one for each channel. ANOVA results showed that the training channel could affect the classification accuracy if SNR = -20dB and 0dB. An overall accuracy of 97.72% has been achieved by one of the models.

Feature Extraction of Surface Electromyography Based on Improved Small-World Leaky Echo State Network

Surface electromyography (sEMG) is an electrophysiological reflection of skeletal muscle contractile activity that can directly reflect neuromuscular activity. It has been a matter of research to investigate feature extraction methods of sEMG signals. In this letter, we propose a feature extraction method of sEMG signals based on the improved small-world leaky echo state network (ISWLESN). The reservoir of leaky echo state network (LESN) is connected by a random network. First, we improved the reservoir of the echo state network (ESN) by these networks and used edge-added probability to improve these networks. That idea enhances the adaptability of the reservoir, the generalization ability, and the stability of ESN. Then we obtained the output weight of the network through training and used it as features. We recorded the sEMG signals during different activities: falling, walking, sitting, squatting, going upstairs, and going downstairs. Afterward, we extracted corresponding features by ISWLESN and used principal component analysis for dimension reduction. At the end, scatter plot, the class separability index, and the Davies-Bouldin index were used to assess the performance of features. The results showed that the ISWLESN clustering performance was better than those of LESN and ESN. By support vector machine, it was also revealed that the performance of ISWLESN for classifying the activities was better than those of ESN and LESN.

Recurrent Neural Network as Estimator for a Virtual sEMG Channel

This study aims at estimating a virtual surface Electromyography (sEMG) channel through a Recurrent Neural Network (RNN) by using Long Short-Term Memory (LSTM) nodes. The virtual channel is used to classify hand postures from the publicly NinaPro database with a multi-class, one-against-all Support Vector Machine (SVM) using the Root Mean Square RMS of the sEMG signal as feature. The classification of the signals through the virtual channel was compared with uncontaminated data and data contaminated with noise saturation. The hit rate from the clean data has averaged 73.96% ± 3.02%. The classification from the contaminated data of one of the channels has improved from 9.29% ± 4.42% to 66.48% ± 6.11% with the virtual channel.

Comparing EMG-Based Human-Machine Interfaces for Estimating Continuous, Coordinated Movements

Electromyography (EMG)-based interfaces are trending toward continuous, simultaneous control with multiple degrees of freedom. Emerging methods range from data-driven approaches to biomechanical model-based methods. However, there has been no direct comparison between these two types of continuous EMG-based interfaces. The aim of this study was to compare a musculoskeletal model (MM) with two data-driven approaches, linear regression (LR) and artificial neural network (ANN), for predicting continuous wrist and hand motions for EMG-based interfaces. Six able-bodied subjects and one transradial amputee subject performed (missing) metacarpophalangeal (MCP) and wrist flexion/extension, simultaneously or independently, while four EMG signals were recorded from forearm muscles. To add variation to the EMG signals, the subjects repeated the MCP and wrist motions at various upper extremity postures. For each subject, the EMG signals collected from the neutral posture were used to build the EMG interfaces; the EMG signals collected from all postures were used to evaluate the interfaces. The performance of the interface was quantified by Pearson's correlation coefficient (r) and the normalized root mean square error (NRMSE) between measured and estimated joint angles. The results demonstrated that the MM predicted movements more accurately, with higher r values and lower NRMSE, than either LR or ANN. Similar results were observed in the transradial amputee. Additionally, the variation in r across postures, an indicator of reliability against posture changes, was significantly lower (better) for the MM than for either LR or ANN. Our findings suggest that incorporating musculoskeletal knowledge into EMG-based human-machine interfaces could improve the estimation of continuous, coordinated motion.

Hands-Free Human-Computer Interface Based on Facial Myoelectric Pattern Recognition

Patients with no or limited hand function usually have difficulty in using conventional input devices such as a mouse or a touch screen. Having the ability of manipulating electronic devices can give patients full access to the digital world, thereby increasing their independence and confidence, and enriching their lives. In this study, a hands-free human-computer interface was developed in order to help patients manipulate computers using facial movements. Five facial movement patterns were detected by four electromyography (EMG) sensors, and classified using myoelectric pattern recognition algorithms. Facial movement patterns were mapped to cursor actions including movements in different directions and click. A typing task and a drawing task were designed in order to assess the interaction performance of the interface in daily use. Ten able-bodied subjects participated in the experiment. In the typing task, the median path efficiency was 80.4%, and the median input rate was 5.9 letters per minute. In the drawing task, the median time to accomplish was 239.9 s. Moreover, all the subjects achieved high classification accuracy (median: 98.0%). The interface driven by facial EMG achieved high performance, and will be assessed on patients with limited hand functions in the future.

Fault-Tolerant Sensor Detection of sEMG signals: Quality Analysis Using a Two-Class Support Vector Machine

The capacity to identify the contamination in surface electromyography (sEMG) signals is necessary for applying the sEMG controlled prosthesis over time. In this paper, the method for the automatic identification of commonly occurring contaminant types in sEMG signals is evaluated. The presented approach uses two-class support vector machine (SVM) trained with clean sEMG and artificially contaminated sEMG. The contaminants considered include electrocardiogram interference, motion artefact, power line interference, amplifier saturation, and electrode displacement. The results demonstrated that the sEMG signal with the contaminants could readily be distinguished, even with increase channels degraded. The SFTD detection depends on the noise type, whether the amputee or non-amputee subjects and which channel is being analysed. This method presented a suitable solution for the detection of contaminants in the sEMG signal, being able to provide the acquired signal validation before the movement intended recognition to operate in an intelligent recognition with greater reliability.

Causes of Performance Degradation in Non-invasive Electromyographic Pattern Recognition in Upper Limb Prostheses

Surface Electromyography (EMG)-based pattern recognition methods have been investigated over the past years as a means of controlling upper limb prostheses. Despite the very good reported performance of myoelectric controlled prosthetic hands in lab conditions, real-time performance in everyday life conditions is not as robust and reliable, explaining the limited clinical use of pattern recognition control. The main reason behind the instability of myoelectric pattern recognition control is that EMG signals are non-stationary in real-life environments and present a lot of variability over time and across subjects, hence affecting the system's performance. This can be the result of one or many combined changes, such as muscle fatigue, electrode displacement, difference in arm posture, user adaptation on the device over time and inter-subject singularity. In this paper an extensive literature review is performed to present the causes of the drift of EMG signals, ways of detecting them and possible techniques to counteract for their effects in the application of upper limb prostheses. The suggested techniques are organized in a table that can be used to recognize possible problems in the clinical application of EMG-based pattern recognition methods for upper limb prosthesis applications and state-of-the-art methods to deal with such problems.

Myoelectric Control Based on A Generic Musculoskeletal Model: Towards A Multi-User Neural-Machine Interface

This study aimed to develop a novel electromyography (EMG)-based neural-machine interface (NMI) that is user-generic for continuously predicting coordinated motion between metacarpophalangeal (MCP) and wrist flexion/extension. The NMI requires a minimum calibration procedure that only involves capturing maximal voluntary muscle contraction for the monitored muscles for individual users. At the center of the NMI is a user-generic musculoskeletal model based on the experimental data collected from 6 able-bodied (AB) subjects and 9 different upper limb postures. The generic model was evaluated on-line on both AB subjects and a transradial amputee. The subjects were instructed to perform a virtual hand/wrist posture matching task with different upper limb postures. The on-line performance of the generic model was also compared with that of the musculoskeletal model customized to each individual user (called "specific model"). All subjects accomplished the assigned virtual tasks while using the user-generic NMI, although the AB subjects produced better performance than the amputee subject. Interestingly, compared to the specific model, the generic model produced comparable completion time, a reduced number of overshoots, and improved path efficiency in the virtual hand/wrist posture matching task. The results suggested that it is possible to design an EMG-driven NMI based on a musculoskeletal model that could fit multiple users, including upper limb amputees, for predicting coordinated MCP and wrist motion. The present new method might address the challenges of existing advanced EMG-based NMI that require frequent and lengthy customization and calibration. Our future research will focus on evaluating the developed NMI for powered prosthetic arms.

Virtual Sensor of Surface Electromyography in a New Extensive Fault-Tolerant Classification System

A few prosthetic control systems in the scientific literature obtain pattern recognition algorithms adapted to changes that occur in the myoelectric signal over time and, frequently, such systems are not natural and intuitive. These are some of the several challenges for myoelectric prostheses for everyday use. The concept of the virtual sensor, which has as its fundamental objective to estimate unavailable measures based on other available measures, is being used in other fields of research. The virtual sensor technique applied to surface electromyography can help to minimize these problems, typically related to the degradation of the myoelectric signal that usually leads to a decrease in the classification accuracy of the movements characterized by computational intelligent systems. This paper presents a virtual sensor in a new extensive fault-tolerant classification system to maintain the classification accuracy after the occurrence of the following contaminants: ECG interference, electrode displacement, movement artifacts, power line interference, and saturation. The Time-Varying Autoregressive Moving Average (TVARMA) and Time-Varying Kalman filter (TVK) models are compared to define the most robust model for the virtual sensor. Results of movement classification were presented comparing the usual classification techniques with the method of the degraded signal replacement and classifier retraining. The experimental results were evaluated for these five noise types in 16 surface electromyography (sEMG) channel degradation case studies. The proposed system without using classifier retraining techniques recovered of mean classification accuracy was of 4% to 38% for electrode displacement, movement artifacts, and saturation noise. The best mean classification considering all signal contaminants and channel combinations evaluated was the classification using the retraining method, replacing the degraded channel by the virtual sensor TVARMA model. This method recovered the classification accuracy after the degradations, reaching an average of 5.7% below the classification of the clean signal, that is the signal without the contaminants or the original signal. Moreover, the proposed intelligent technique minimizes the impact of the motion classification caused by signal contamination related to degrading events over time. There are improvements in the virtual sensor model and in the algorithm optimization that need further development to provide an increase the clinical application of myoelectric prostheses but already presents robust results to enable research with virtual sensors on biological signs with stochastic behavior.

Anomaly Detection of Electromyographic Signals

In this paper, we provide a robust framework to detect anomalous electromyographic (EMG) signals and identify contamination types. As a first step for feature selection, optimally selected Lawton wavelets transform is applied. Robust principal component analysis (rPCA) is then performed on these wavelet coefficients to obtain features in a lower dimension. The rPCA based features are used for constructing a self-organizing map (SOM). Finally, hierarchical clustering is applied on the SOM that separates anomalous signals residing in the smaller clusters and breaks them into logical units for contamination identification. The proposed methodology is tested using synthetic and real world EMG signals. The synthetic EMG signals are generated using a heteroscedastic process mimicking desired experimental setups. A sub-part of these synthetic signals is introduced with anomalies. These results are followed with real EMG signals introduced with synthetic anomalies. Finally, a heterogeneous real world data set is used with known quality issues under an unsupervised setting. The framework provides recall of 90% (± 3.3) and precision of 99%(±0.4).

Limb Position Tolerant Pattern Recognition for Myoelectric Prosthesis Control with Adaptive Sparse Representations From Extreme Learning

Myoelectric signals can be used to predict the intended movements of an amputee for prosthesis control. However, untrained effects like limb position changes influence myoelectric signal characteristics, hindering the ability of pattern recognition algorithms to discriminate among motion classes. Despite frequent and long training sessions, these deleterious conditional influences may result in poor performance and device abandonment.

GOAL

We present a robust sparsity-based adaptive classification method that is significantly less sensitive to signal deviations resulting from untrained conditions.

METHODS

We compare this approach in the offline and online contexts of untrained upper-limb positions for amputee and able-bodied subjects to demonstrate its robustness compared against other myoelectric classification methods.

RESULTS

We report significant performance improvements () in untrained limb positions across all subject groups.

SIGNIFICANCE

The robustness of our suggested approach helps to ensure better untrained condition performance from fewer training conditions.

CONCLUSIONS

This method of prosthesis control has the potential to deliver real-world clinical benefits to amputees: better condition-tolerant performance, reduced training burden in terms of frequency and duration, and increased adoption of myoelectric prostheses.

Evaluation of EMG pattern recognition for upper limb prosthesis control: a case study in comparison with direct myoelectric control

Background

Although electromyogram (EMG) pattern recognition (PR) for multifunctional upper limb prosthesis control has been reported for decades, the clinical benefits have rarely been examined. The study purposes were to: 1) compare self-report and performance outcomes of a transradial amputee immediately after training and one week after training of direct myoelectric control and EMG pattern recognition (PR) for a two-degree-of-freedom (DOF) prosthesis, and 2) examine the change in outcomes one week after pattern recognition training and the rate of skill acquisition in two subjects with transradial amputations.

Methods

In this cross-over study, participants were randomized to receive either PR control or direct control (DC) training of a 2 DOF myoelectric prosthesis first. Participants were 2 persons with traumatic transradial (TR) amputations who were 1 DOF myoelectric users. Outcomes, including measures of dexterity with and without cognitive load, activity performance, self-reported function, and prosthetic satisfaction were administered immediately and 1 week after training. Speed of skill acquisition was assessed hourly. One subject completed training under both PR control and DC conditions. Both subjects completed PR training and testing. Outcomes of test metrics were analyzed descriptively.

Results

Comparison of the two control strategies in one subject who completed training in both conditions showed better scores in 2 (18%) dexterity measures, 1 (50%) dexterity measure with cognitive load, and 1 (50%) self-report functional measure using DC, as compared to PR. Scores of all other metrics were comparable. Both subjects showed decline in dexterity after training. Findings related to rate of skill acquisition varied considerably by subject.

Conclusions

Outcomes of PR and DC for operating a 2-DOF prosthesis in a single subject cross-over study were similar for 74% of metrics, and favored DC in 26% of metrics. The two subjects who completed PR training showed decline in dexterity one week after training ended. Findings related to rate of skill acquisition varied considerably by subject. This study, despite its small sample size, highlights a need for additional research quantifying the functional and clinical benefits of PR control for upper limb prostheses.

Electronic supplementary material

The online version of this article (10.1186/s12984-018-0361-3) contains supplementary material, which is available to authorized users.

Electrode-shift Tolerant Myoelectric Movement-pattern Classification using Extreme Learning for Adaptive Sparse Representations

Myoelectric signal patterns can be used to predict the intended movements of amputees for prosthesis activation. Real-world prosthesis use introduces a variety of unpredictable conditional influences on these patterns, hindering the performance of classification algorithms and potentially leading to device abandonment. We have discovered a state-of-the-art classification method which is significantly more tolerant to these conditional influences. In our prior work, we presented a robust sparsity-based adaptive classification method that is tolerant to pattern deviations resulting from untrained limb positions and the prosthesis load. Herein, we demonstrate that this method is tolerant to the shifting or misalignment of the contact-electrode array which occurs during prosthesis use. We demonstrate the robustness of this approach in untrained electrode-site locations for amputee and able-bodied subjects, and report significant performance improvements over conventional myoelectric pattern recognition approaches. By showing that a single, unified method is robust across a variety of real-world condition spaces, clinicians are more likely to incorporate this method into myoelectric prosthesis controllers, resulting in improved utility and increased adoption among amputee users.

Improved prosthetic hand control with concurrent use of myoelectric and inertial measurements

Background

Myoelectric pattern recognition systems can decode movement intention to drive upper-limb prostheses. Despite recent advances in academic research, the commercial adoption of such systems remains low. This limitation is mainly due to the lack of classification robustness and a simultaneous requirement for a large number of electromyogram (EMG) electrodes. We propose to address these two issues by using a multi-modal approach which combines surface electromyography (sEMG) with inertial measurements (IMs) and an appropriate training data collection paradigm. We demonstrate that this can significantly improve classification performance as compared to conventional techniques exclusively based on sEMG signals.

Methods

We collected and analyzed a large dataset comprising recordings with 20 able-bodied and two amputee participants executing 40 movements. Additionally, we conducted a novel real-time prosthetic hand control experiment with 11 able-bodied subjects and an amputee by using a state-of-the-art commercial prosthetic hand. A systematic performance comparison was carried out to investigate the potential benefit of incorporating IMs in prosthetic hand control.

Results

The inclusion of IM data improved performance significantly, by increasing classification accuracy (CA) in the offline analysis and improving completion rates (CRs) in the real-time experiment. Our findings were consistent across able-bodied and amputee subjects. Integrating the sEMG electrodes and IM sensors within a single sensor package enabled us to achieve high-level performance by using on average 4-6 sensors.

Conclusions

The results from our experiments suggest that IMs can form an excellent complimentary source signal for upper-limb myoelectric prostheses. We trust that multi-modal control solutions have the potential of improving the usability of upper-extremity prostheses in real-life applications.

Electronic supplementary material

The online version of this article (doi:10.1186/s12984-017-0284-4) contains supplementary material, which is available to authorized users.

Classification complexity in myoelectric pattern recognition

Background

Limb prosthetics, exoskeletons, and neurorehabilitation devices can be intuitively controlled using myoelectric pattern recognition (MPR) to decode the subject’s intended movement. In conventional MPR, descriptive electromyography (EMG) features representing the intended movement are fed into a classification algorithm. The separability of the different movements in the feature space significantly affects the classification complexity. Classification complexity estimating algorithms (CCEAs) were studied in this work in order to improve feature selection, predict MPR performance, and inform on faulty data acquisition.

Methods

CCEAs such as nearest neighbor separability (NNS), purity, repeatability index (RI), and separability index (SI) were evaluated based on their correlation with classification accuracy, as well as on their suitability to produce highly performing EMG feature sets. SI was evaluated using Mahalanobis distance, Bhattacharyya distance, Hellinger distance, Kullback–Leibler divergence, and a modified version of Mahalanobis distance. Three commonly used classifiers in MPR were used to compute classification accuracy (linear discriminant analysis (LDA), multi-layer perceptron (MLP), and support vector machine (SVM)). The algorithms and analytic graphical user interfaces produced in this work are freely available in BioPatRec.

Results

NNS and SI were found to be highly correlated with classification accuracy (correlations up to 0.98 for both algorithms) and capable of yielding highly descriptive feature sets. Additionally, the experiments revealed how the level of correlation between the inputs of the classifiers influences classification accuracy, and emphasizes the classifiers’ sensitivity to such redundancy.

Conclusions

This study deepens the understanding of the classification complexity in prediction of motor volition based on myoelectric information. It also provides researchers with tools to analyze myoelectric recordings in order to improve classification performance.

Using non-iterative methods and random weight networks to classify upper-limb movements through sEMG signals

This paper presents the use of two non-iterative methods to perform the classification of 17 different upper-limb movements through sEMG signal processing. The two methods were compared with a SVM classifier using three different databases involving amputee subjects. The non-iterative methods presented equivalent or superior classification accuracy than SVM method. Thereafter a stage of PCA pre-processing method was used in order to promote a better class separation prior the non-iterative classifiers. The best accuracy result without PCA was achieved by the Regularized Extreme Learning Machines algorithm (88,4% for non-amputee subjects and 79,4% for the amputee). The PCA method used boosted the accuracy of the two non-iterative methods which the mean accuracy was 94% for the non-amputee subjects and 85% for the amputee subjects.

Lumped-parameter electromyogram-driven musculoskeletal hand model: A potential platform for real-time prosthesis control

Simple, lumped-parameter musculoskeletal models may be more adaptable and practical for clinical real-time control applications, such as prosthesis control. In this study, we determined whether a lumped-parameter, EMG-driven musculoskeletal model with four muscles could predict wrist and metacarpophalangeal (MCP) joint flexion/extension. Forearm EMG signals and joint kinematics were collected simultaneously from 5 able-bodied (AB) subjects. For one subject with unilateral transradial amputation (TRA), joint kinematics were collected from the sound arm during bilateral mirrored motion. Twenty-two model parameters were optimized such that joint kinematics predicted by EMG-driven forward dynamic simulation closely matched measured kinematics. Cross validation was employed to evaluate the model kinematic predictions using Pearson׳s correlation coefficient (r). Model predictions of joint angles were highly to very highly positively correlated with measured values at the wrist (AB mean r=0.94, TRA r=0.92) and MCP (AB mean r=0.88, TRA r=0.93) joints during single-joint wrist and MCP movements, respectively. In simultaneous multi-joint movement, the prediction accuracy for TRA at the MCP joint decreased (r=0.56), while r-values derived from AB subjects and TRA wrist motion were still above 0.75. Though parameters were optimized to match experimental sub-maximal kinematics, passive and maximum isometric joint moments predicted by the model were comparable to reported experimental measures. Our results showed the promise of a lumped-parameter musculoskeletal model for hand/wrist kinematic estimation. Therefore, the model might be useful for EMG control of powered upper limb prostheses, but more work is needed to demonstrate its online performance.