A Classification Method for Myoelectric Control of Hand Prostheses Inspired by Muscle Coordination

Adaptive neuro-fuzzy inference system model driven by the non-negative matrix factorization-extracted muscle synergy patterns to estimate lower limb joint movements

BACKGROUND AND OBJECTIVE

For patients with movement disorders, the main clinical focus is on exercise rehabilitation to help recover lost motor function, which is achieved by relevant assisted equipment. The basis for seamless control of the assisted equipment is to achieve accurate inference of the user's movement intentions in the human-machine interface. This study proposed a novel movement intention detection technology for estimating lower limb joint continuous kinematic variables following muscle synergy patterns, to develop applications for more efficient assisted rehabilitation training.

METHODS

This study recruited 16 healthy males and 16 male patients with symptomatic patellar tendinopathy (VISA-P: 59.1 ± 8.7). The surface electromyography of 12 muscles and lower limb joint kinematic and kinetic data from healthy subjects and patients during step-off landings from 30 cm-high stair steps were collected. We subsequently solved the preprocessed data based on the established recursive model of second-order differential equation to obtain the muscle activation matrix, and then imported it into the non-negative matrix factorization model to obtain the muscle synergy matrix. Finally, the lower limb neuromuscular synergy pattern was then imported into the developed adaptive neuro-fuzzy inference system non-linear regression model to estimate the human movement intention during this movement pattern.

RESULTS

Six muscle synergies were determined to construct the muscle synergy pattern driven ANFIS model. Three fuzzy rules were determined in most estimation cases. Combining the results of the four error indicators across the estimated variables indicates that the current model has excellent estimated performance in estimating lower limb joint movement. The estimation errors between the healthy (Angle: R2=0.98±0.03; Torque: R2=0.96±0.04) and patient (Angle: R2=0.98±0.02; Torque: R2=0.96±0.03) groups are consistent.

CONCLUSION

The proposed model of this study can accurately and reliably estimate lower limb joint movements, and the effectiveness will also be radiated to the patient group. This revealed that our models also have certain advantages in the recognition of motor intentions in patients with relevant movement disorders. Future work from this study can be focused on sports rehabilitation in the clinical field by achieving more flexible and precise movement control of the lower limb assisted equipment to help the rehabilitation of patients.

A Novel Screen-Printed Textile Interface for High-Density Electromyography Recording

Recording electrical muscle activity using a dense matrix of detection points (high-density electromyography, EMG) is of interest in a range of different applications, from human-machine interfacing to rehabilitation and clinical assessment. The wider application of high-density EMG is, however, limited as the clinical interfaces are not convenient for practical use (e.g., require conductive gel/cream). In the present study, we describe a novel dry electrode (TEX) in which the matrix of sensing pads is screen printed on textile and then coated with a soft polymer to ensure good skin-electrode contact. To benchmark the novel solution, an identical electrode was produced using state-of-the-art technology (polyethylene terephthalate with hydrogel, PET) and a process that ensured a high-quality sample. The two electrodes were then compared in terms of signal quality as well as functional application. The tests showed that the signals collected using PET and TEX were characterised by similar spectra, magnitude, spatial distribution and signal-to-noise ratio. The electrodes were used by seven healthy subjects and an amputee participant to recognise seven hand gestures, leading to similar performance during offline analysis and online control. The comprehensive assessment, therefore, demonstrated that the proposed textile interface is an attractive solution for practical applications.

Reducing the number of EMG electrodes during online hand gesture classification with changing wrist positions

Background

Myoelectric control based on hand gesture classification can be used for effective, contactless human–machine interfacing in general applications (e.g., consumer market) as well as in the clinical context. However, the accuracy of hand gesture classification can be impacted by several factors including changing wrist position. The present study aimed at investigating how channel configuration (number and placement of electrode pads) affects performance in hand gesture recognition across wrist positions, with the overall goal of reducing the number of channels without the loss of performance with respect to the benchmark (all channels).

Methods

Matrix electrodes (256 channels) were used to record high-density EMG from the forearm of 13 healthy subjects performing a set of 8 gestures in 3 wrist positions and 2 force levels (low and moderate). A reduced set of channels was chosen by applying sequential forward selection (SFS) and simple circumferential placement (CIRC) and used for gesture classification with linear discriminant analysis. The classification success rate and task completion rate were the main outcome measures for offline analysis across the different number of channels and online control using 8 selected channels, respectively.

Results

The offline analysis demonstrated that good accuracy (> 90%) can be achieved with only a few channels. However, using data from all wrist positions required more channels to reach the same performance. Despite the targeted placement (SFS) performing similarly to CIRC in the offline analysis, the task completion rate [median (lower–upper quartile)] in the online control was significantly higher for SFS [71.4% (64.8–76.2%)] compared to CIRC [57.1% (51.8–64.8%), p < 0.01], especially for low contraction levels [76.2% (66.7–84.5%) for SFS vs. 57.1% (47.6–60.7%) for CIRC, p < 0.01]. For the reduced number of electrodes, the performance with SFS was comparable to that obtained when using the full matrix, while the selected electrodes were highly subject-specific.

Conclusions

The present study demonstrated that the number of channels required for gesture classification with changing wrist positions could be decreased substantially without loss of performance, if those channels are placed strategically along the forearm and individually for each subject. The results also emphasize the importance of online assessment and motivate the development of configurable matrix electrodes with integrated channel selection.

Evaluation of Synergy-Based Hand Gesture Recognition Method Against Force Variation for Robust Myoelectric Control

The non-stationary characteristics of surface electromyography (sEMG) and possible adverse variations in real-world conditions make it still an open challenge to realize robust myoelectric control (MEC) for multifunctional prostheses. Variable muscle contraction level is one of the handicaps that may degrade the performance of MEC. In this study, we proposed a force-invariant intent recognition method based on muscle synergy analysis (MSA) in the setting of three self-defined force levels (low, medium, and high). Specifically, a fast matrix factorization algorithm based on alternating non-negativity constrained least squares (NMF/ANLS) was chosen to extract task-specific synergies associated with each of six hand gestures in the training stage; while for the testing samples, we used the non-negative least square (NNLS) method to estimate neural commands for movement classification. The performance of proposed method was compared with conventional pattern recognition (PR) method consisting of LDA (linear discrimination analysis) classifier and representative features in three offline evaluation scenarios. Statistical tests on ten able-bodied subjects revealed no significant difference in intra-force-level (p = 0.353) and multi-force-level (p = 0.695) accuracy; But the synergy-based method performed significantly better than conventional PR-based method under inter-force-level conditions (p < 0.05). Similar results were observed for nine amputee subjects though there was a drop in the classification accuracy. This study was the first to concurrently demonstrate the robustness and predictive power of task-specific synergies under variant force levels and explore their potential for reliable intent recognition against force variation. Although the online performance is yet to be demonstrated, the proposed method is characterized by simple training procedure and acceptable computational efficiency, which would potentially provide an alternative approach for the development of clinically viable prostheses and rehabilitation robots driven by sEMG.

A novel framework for designing a multi-DoF prosthetic wrist control using machine learning

Prosthetic arms can significantly increase the upper limb function of individuals with upper limb loss, however despite the development of various multi-DoF prosthetic arms the rate of prosthesis abandonment is still high. One of the major challenges is to design a multi-DoF controller that has high precision, robustness, and intuitiveness for daily use. The present study demonstrates a novel framework for developing a controller leveraging machine learning algorithms and movement synergies to implement natural control of a 2-DoF prosthetic wrist for activities of daily living (ADL). The data was collected during ADL tasks of ten individuals with a wrist brace emulating the absence of wrist function. Using this data, the neural network classifies the movement and then random forest regression computes the desired velocity of the prosthetic wrist. The models were trained/tested with ADLs where their robustness was tested using cross-validation and holdout data sets. The proposed framework demonstrated high accuracy (F-1 score of 99% for the classifier and Pearson's correlation of 0.98 for the regression). Additionally, the interpretable nature of random forest regression was used to verify the targeted movement synergies. The present work provides a novel and effective framework to develop an intuitive control for multi-DoF prosthetic devices.

Evaluating User and Machine Learning in Short- and Long-Term Pattern Recognition-Based Myoelectric Control

Proper training is essential to achieve reliable pattern recognition (PR) based myoelectric control. The amount of training is commonly determined by experience. The purpose of this study is to provide an offline validation method that makes the offline performance transferable to online control and find the proper amount of training that achieves good online performance. In the offline experiment, eight able-bodied subjects and three amputees participated in a ten-day training. Repeatability index (RI) and classification error (CE) were used to evaluate user learning and machine learning, respectively. The performance of cross-validation (CV) and time serial related validation (TSV) was compared. Learning curves were established with different training trials by TSV. In the online experiment, sixteen able-bodied subjects were randomly divided into two groups with one- or five-trial training, respectively, followed by participating in the test with and without classifier-output feedback. The correlation between offline and online tests was analyzed. Results indicated that five-trial training was proper to train the user and the classifier. The long-term retention of skills could not shorten the learning process. The correlation between CEs of TSV and the online test was strong ( r=0.87 ) with five-trial training, while the correlation between CEs of CV and the online test was weak ( r=0.30 ). Outcomes demonstrate that offline performance evaluated by TSV is transferable to online performance and the learning process can guide the user to achieve good online myoelectric control with minimum training.

Activities of daily living with bionic arm improved by combination training and latching filter in prosthesis control comparison

BACKGROUND

Advanced prostheses can restore function and improve quality of life for individuals with amputations. Unfortunately, most commercial control strategies do not fully utilize the rich control information from residual nerves and musculature. Continuous decoders can provide more intuitive prosthesis control using multi-channel neural or electromyographic recordings. Three components influence continuous decoder performance: the data used to train the algorithm, the algorithm, and smoothing filters on the algorithm's output. Individual groups often focus on a single decoder, so very few studies compare different decoders using otherwise similar experimental conditions.

METHODS

We completed a two-phase, head-to-head comparison of 12 continuous decoders using activities of daily living. In phase one, we compared two training types and a smoothing filter with three algorithms (modified Kalman filter, multi-layer perceptron, and convolutional neural network) in a clothespin relocation task. We compared training types that included only individual digit and wrist movements vs. combination movements (e.g., simultaneous grasp and wrist flexion). We also compared raw vs. nonlinearly smoothed algorithm outputs. In phase two, we compared the three algorithms in fragile egg, zipping, pouring, and folding tasks using the combination training and smoothing found beneficial in phase one. In both phases, we collected objective, performance-based (e.g., success rate), and subjective, user-focused (e.g., preference) measures.

RESULTS

Phase one showed that combination training improved prosthesis control accuracy and speed, and that the nonlinear smoothing improved accuracy but generally reduced speed. Phase one importantly showed simultaneous movements were used in the task, and that the modified Kalman filter and multi-layer perceptron predicted more simultaneous movements than the convolutional neural network. In phase two, user-focused metrics favored the convolutional neural network and modified Kalman filter, whereas performance-based metrics were generally similar among all algorithms.

CONCLUSIONS

These results confirm that state-of-the-art algorithms, whether linear or nonlinear in nature, functionally benefit from training on more complex data and from output smoothing. These studies will be used to select a decoder for a long-term take-home trial with implanted neuromyoelectric devices. Overall, clinical considerations may favor the mKF as it is similar in performance, faster to train, and computationally less expensive than neural networks.

Robustness of Muscle Synergies under Variant Muscle Contraction Force during Forearm Movements

The Electromyography-based Pattern-Recognition (EMG-PR) framework has been investigated for almost three decades towards developing an intuitive myoelectric prosthesis. To utilize the knowledge of the underlying neurophysiological processes of natural movements, the concept of muscle synergy has been applied in prosthesis control and proved to be of great potential recently. For a muscle-synergy-based myoelectric system, the variation of muscle contraction force is also a confounding factor. This study evaluates the robustness of muscle synergies under a variant force level for forearm movements. Six channels of forearm surface EMG were recorded from six healthy subjects when they performed 4 movements (hand open, hand close, wrist flexion, and wrist extension) using low, moderate, and high force, respectively. Muscle synergies were extracted from the EMG using the alternating nonnegativity constrained least squares and active set (NNLS) algorithm. Three analytic strategies were adopted to examine whether muscle synergies were conserved under different force levels. Our results consistently showed that there exists fixed, robust muscle synergies across force levels. This outcome would provide valuable insights to the implementation of muscle- synergy-based assistive technology for the upper extremity.

Adaptive Spatial Filtering of High-Density EMG for Reducing the Influence of Noise and Artefacts in Myoelectric Control

Electromyography (EMG) is a source of neural information for controlling neuroprosthetic devices. To enhance the information content of conventional bipolar EMG, high-density EMG systems include tens to hundreds of closely spaced electrodes that non-invasively record the electrical activity of muscles with high spatial resolution. Despite the advantages of relying on multiple signal sources, however, variations in electrode-skin contact impedance and noise remain challenging for multichannel myocontrol systems. These spatial and temporal non-stationarities negatively impact the control accuracy and therefore substantially limit the clinical viability of high-density EMG techniques. Here, we propose an adaptive algorithm for automatic artefact/noise detection and attenuation for high-density EMG control. The method infers the presence of noise in each EMG channel by spectro-temporal measures of signal similarity. These measures are then used for establishing a scoring system based on an adaptive weighting and reinforcement formulation. The method was experimentally tested as a pre-processing step for a multi-class discrimination problem of 4-digit activation. The approach was proven to enhance the discriminative information content of high-density EMG signals, as well as to attenuate non-stationary artefacts, with improvements in accuracy and robustness of the classification.

Modulation of muscle synergies for multiple forearm movements under variant force and arm position constraints

OBJECTIVE

To promote clinical applications of muscle-synergy-based neurorehabilitation techniques, this study aims to clarify any potential modulations of both the muscular compositions and temporal activations of forearm muscle synergies for multiple movements under variant force levels and arm positions.

APPROACH

Two groups of healthy subjects participated in this study. Electromyography (EMG) signals were collected when they performed four hand and wrist movements under variant constraints-three different force levels for one group and five arm positions for the other. Muscle synergies were extracted from the EMGs, and their robustness across variant force levels and arm positions was separately assessed by evaluating their across-condition structure similarity, cross-validation, and cluster analysis. The synergies' activation coefficients across the variant constraints were also compared, and the coefficients were used to discriminate the different force levels and the arm positions, respectively.

MAIN RESULTS

Overall, the muscle synergies were relatively fixed across variant constraints, but they were more robust to variant forces than to changing arm positions. The activations of muscle synergies depended largely on the level of contraction force and could discriminate the force levels very well, but the coefficients corresponding to different arm positions discriminated the positions with lower accuracy. Similar results were found for all types of forearm movement analyzed.

SIGNIFICANCE

With our experiment and subject-specific analysis, only slight modulation of the muscular compositions of forearm muscle synergies was found under variant force and arm position constraints. Our results may shed valuable insights to future design of both muscle-synergy-based assistive robots and motor-function assessments.

An Online Spectral Information-Enhanced Approach for Artifact Detection and Fault Attentuation in Myoelectric Control

In myocontrol of neuroprosthetic devices, multichannel electromyography (EMG) can be used to decode the intended motor command, based on distributed activation patterns of stump muscles. In this regard, the high density EMG (HD-EMG) approach allows for enhancement of the spatiotemporal resolution for motor intention detection. Despite the advantages of relying on several EMG channels, the challenge of high-density electrode systems is the dynamically changing electrode-skin contact impedance, which can affect a considerable number of electrodes over the time of data acquisition. This can result in obtaining unreliable, low-quality EMG recording with a distributed artifact pattern over the grid of EMG sensors. To address this issue, we propose a novel online approach for adaptive information extraction and enhancement for automatic artifact detection and attenuation in HD-EMG-based myocontrol of prosthetic devices. The method is based on an adaptive weighting scheme that modifies the contribution of each HD-EMG channel considering the spectral information content relative to artifacts. The technique (named IE-HD-EMG) was tested as an online pre-conditioning step for a challenging multiclass classification problem of 4-finger activation, using linear discriminant analysis. It is shown that for this application, the proposed IE-HD-EMG technique led to a superior performance in finger activation recognition (79.25% accuracy, 89% sensitivity, 89.15% specificity) in comparison to the conventional HD-EMG recording under the same condition without the proposed approach (56.25% accuracy, 61.3% sensitivity, 67% specificity). Therefore, the proposed technique can have a significant potential to expand the clinical viability of HD-EMG systems.