A Hand Gesture Recognition Strategy Based on Virtual-Dimension Increase of EMG.

The electromyography(EMG) signal is the biocurrent associated with muscle contraction and can be used as the input signal to a myoelectric intelligent bionic hand to control different gestures of the hand. Increasing the number of myoelectric-signal channels can yield richer information of motion intention and improve the accuracy of gesture recognition. However, as the number of acquisition channels increases, its effect on the improvement of the accuracy of gesture recognition gradually diminishes, resulting in the improvement of the control effect reaching a plateau. To address these problems, this paper presents a proposed method to improve gesture recognition accuracy by virtually increasing the number of EMG signal channels. This method is able to improve the recognition accuracy of various gestures by virtually increasing the number of EMG signal channels and enriching the motion intention information extracted from data collected from a certain number of physical channels, ultimately providing a solution to the issue of the recognition accuracy plateau caused by saturation of information from physical recordings. Meanwhile, based on the idea of the filtered feature selection method, a quantitative measure of sample sets (separability of feature vectors [SFV]) derived from the divergence and correlation of the extracted features is introduced. The SFV value can predict the classification effect before performing the classification, and the effectiveness of the virtual-dimension increase strategy is verified from the perspective of feature set differentiability change. Compared to the statistical motion intention recognition success rate, SFV is a more representative and faster measure of classification effectiveness and is also suitable for small sample sets.

Novel prosthesis with proximity sensor for upper extremity phocomelia.

We herein report on the application of a novel motorized prosthetic hand in a child with upper extremity phocomelia.

Comparison of Bipolar and Monopolar Electrode Configurations for FES on Biceps Brachii.

Function electrical stimulation (FES) is recommended as one of the effective methods for rehabilitation of motor function after stroke. There are two forms to deliver electrical stimulation to induce muscle contraction: Bipolar electrode configuration with two electrodes of the same size, and monopolar electrode configuration with a bigger electrode as an indifferent electrode and a smaller one as an active electrode. The purpose of this study is to compare the two kinds of configuration on biceps brachii in terms of induced muscle contraction force and muscle fatigue. In the experiment, electrical stimulation was applied on biceps brachii muscles of the right arm. Isometric contraction was induced by fixing the elbow joint during the stimulation. The experimental results showed that the induced contraction force was bigger using monopolar electrode configuration with the indifferent electrode on the antagonist muscle, and there was no significant difference in muscle fatigue between the configurations. Monopolar electrode configuration with the indifferent electrode on the antagonist muscle was suggested as the most effective method for FES on biceps brachii.Clinical Relevance- This study establishes an effective electrode configuration for FES on biceps brachii.

Highly Anthropomorphic Finger Design With a Novel Friction Clutch for Achieving Human-Like Reach-and-Grasp Movements.

In the design of prosthetic hand fingers, achieving human-like movement while meeting anthropomorphic demands such as appearance, size, and lightweight is quite challenging. Human finger movement involves two distinct motion characters during natural reach-and-grasp tasks: consistency in the reaching stage and adaptability in the grasping stage. The former one enhances grasp stability and reduces control complexity; the latter one promotes the adaptability of finger to various objects. However, conventional tendon-driven prosthetic finger designs typically incorporate bulky actuation modules or complex tendon routes to reconcile the consistency and adaptability. In contrast, we propose a novel friction clutch consisting of a single tendon and slider, which is simple and compact enough to be configurated within the metacarpal bone. Through tactfully exploiting the friction force to balance the gravity effect on each phalanx during finger motion, this design effectively combines both consistency and adaptability. As a result, the prosthetic finger can maintain consistent motion unaffected by any spatial posture during reaching, execute adaptive motion during grasping, and automatically switch between them, resulting in human-like reach-and-grasp movements. Additionally, the proposed finger achieves a highly anthropomorphic design, weighing only 18.9 g and possessing the same size as an adult's middle finger. Finally, a series of experiments validate the theoretical effectiveness and motion performance of the proposed design. Remarkably, the mechanical principle of the proposed friction clutch is beneficial to achieve highly anthropomorphic design, providing not only a new strategy to prosthetic hand design but also great potential in hand rehabilitation.

High-precision dynamic torque control of high stiffness actuator for humanoids.

The high stiffness actuator (HSA), applied to each joint of an electrical driven humanoid robot, can directly affect the motion performance of the torque-controlled humanoid robots. For high control performance of HSA, a high-precision dynamic torque control (HDTC) is proposed. The HDTC consists of two phases: (1) A novel dynamic current control is used to linearize high stiffness actuator torque control system, which can estimate and compensate the nonlinear coupling parts; (2) An enhanced internal model control is designed to ensure high tracking accuracy in the system containing noisy torque signal and even numerical differentiation signals. Benefitting from dynamic current control and the enhanced internal model control, the proposed HDTC is accurate and adaptable. Finally, the superiority of the HDTC is verified with comparative experiments.

Improvement of Precision Grasping Performance by Interaction Between Soft Finger Pulp and Hard Nail.

In this study, we investigated the effect of the presence or absence of fingernails on precision grasping using artificial anthropomimetic fingers. We hypothesized that fingernails improve precision grasping performance by increasing the friction coefficient while suppressing fingertip deformation. To test our hypothesis, we developed artificial fingertips, each composed of bone, nail, skin, and soft tissue, and fabricated three types of artificial fingers with different skin softness grades and artificial fingers without nails as the control condition. Pullout experiments of cylindrical objects and T-shaped blocks were conducted using the developed artificial fingertips with and without nails, and the magnitude of the holding force was compared. The nail contributed to object grasping stability because the magnitude of the holding force was significantly increased by the presence of the nail in the artificial fingertip with soft skin. The rate of increase in the magnitude of the holding force of the T-shaped block was more significant (3.10 times maximum) compared with the cylindrical object (1.08 times maximum) because the finger pulp deformation was suppressed by the nail, and the form closure, that is, geometric constraint, was formed for the grasping object. The results of this study show that soft fingertips and hard nails can significantly improve the grasping performance of soft robotic hands. And these results suggest that the human nail improves precision grasping performance by forming geometric constraints on the grasped object, suppressing finger pulp deformation.

Towards Optimizing the Quality of Long-Term Physiological Signals Monitoring by Using Anhydrous Carbon Paste Electrode.

Long-term physiological signal monitoring is very important for the diagnosis of health conditions that occur randomly and cannot be easily detected by a short period of a hospital visit. However, the conventional wet electrodes suffered from the problem of signal quality degradation due to the gradual dehydration of the conductive gel. An anhydrous carbon paste electrode (CPE) constructed by a composite of carbon black and polydimethylsiloxane was proposed to enable long-term physiological signal monitoring without signal quality degradation as time elapses. The performance was systematically compared with conventional electrodes when measuring long-term physiological signals including electrocardiogram (ECG), electromyogram (EMG), electroencephalogram (EEG) and auditory brainstem response (ABR). The proposed CPE showed more stable skin-electrode impedance and higher signal qualities as the monitoring time increased up to 48 days, with signal-to-noise ratios (SNRs) of 16.43 ± 10.39 dB higher for ECG and 24.30 ± 7.79 dB higher for EMG when compared with wet electrodes. The CPE method could also obtain more consistent ABR waveform morphologies and could measure EEG under sweating conditions. It is believed that the proposed CPE could be a potential candidate for durable and robust wearable sensors systems on long-term physiological signal monitoring.

Coevolution of Myoelectric Hand Control under the Tactile Interaction among Fingers and Objects.

The usability of a prosthetic hand differs significantly from that of a real hand. Moreover, the complexity of manipulation increases as the number of degrees of freedom to be controlled increases, making manipulation with biological signals extremely difficult. To overcome this problem, users need to select a grasping posture that is adaptive to the object and a stable grasping method that prevents the object from falling. In previous studies, these have been left to the operating skills of the user, which is extremely difficult to achieve. In this study, we demonstrate how stable and adaptive grasping can be achieved according to the object regardless of the user's operation technique. The required grasping technique is achieved by determining the correlation between the motor output and each sensor through the interaction between the prosthetic hand and the surrounding stimuli, such as myoelectricity, sense of touch, and grasping objects. The agents of the 16-DOF robot hand were trained with the myoelectric signals of six participants, including one child with a congenital forearm deficiency. Consequently, each agent could open and close the hand in response to the myoelectric stimuli and could accomplish the object pickup task. For the tasks, the agents successfully identified grasping patterns suitable for practical and stable positioning of the objects. In addition, the agents were able to pick up the object in a similar posture regardless of the participant, suggesting that the hand was optimized by evolutionary computation to a posture that prevents the object from being dropped.

Towards high-accuracy classifying attention-deficit/hyperactivity disorders using CNN-LSTM model.

Objective. The neurocognitive attention functions involve the cooperation of multiple brain regions, and the defects in the cooperation will lead to attention-deficit/hyperactivity disorder (ADHD), which is one of the most common neuropsychiatric disorders for children. The current ADHD diagnosis is mainly based on a subjective evaluation that is easily biased by the experience of the clinicians and lacks the support of objective indicators. The purpose of this study is to propose a method that can effectively identify children with ADHD.Approach. In this study, we proposed a CNN-LSTM model to solve the three-class problems of classifying ADHD, attention deficit disorder (ADD) and healthy children, based on a public electroencephalogram (EEG) dataset that includes event-related potential (ERP) EEG signals of 144 children. The convolution visualization and saliency map methods were used to observe the features automatically extracted by the proposed model, which could intuitively explain how the model distinguished different groups.Main results. The results showed that our CNN-LSTM model could achieve an accuracy as high as 98.23% in a five-fold cross-validation method, which was significantly better than the current state-of-the-art CNN models. The features extracted by the proposed model were mainly located in the frontal and central areas, with significant differences in the time period mappings among the three different groups. The P300 and contingent negative variation (CNV) in the frontal lobe had the largest decrease in the healthy control (HC) group, and the ADD group had the smallest decrease. In the central area, only the HC group had a significant negative oscillation of CNV waves.Significance. The results of this study suggest that the CNN-LSTM model can effectively identify children with ADHD and its subtypes. The visualized features automatically extracted by this model could better explain the differences in the ERP response among different groups, which is more convincing than previous studies, and it could be used as more reliable neural biomarkers to help with more accurate diagnosis in the clinics.

Design of an Effective Prosthetic Hand System for Adaptive Grasping with the Control of Myoelectric Pattern Recognition Approach.

In this paper, we develop a prosthetic bionic hand system to realize adaptive gripping with two closed-loop control loops by using a linear discriminant analysis algorithm (LDA). The prosthetic hand contains five fingers and each finger is driven by a linear servo motor. When grasping objects, four fingers except the thumb would adjust automatically and bend with an appropriate gesture, while the thumb is stretched and bent by the linear servo motor. Since the change of the surface electromechanical signal (sEMG) occurs before human movement, the recognition of sEMG signal with LDA algorithm can help to obtain people's action intention in advance, and then timely send control instructions to assist people to grasp. For activity intention recognition, we extract three features, Variance (VAR), Root Mean Square (RMS) and Minimum (MIN) for recognition. As the results show, it can achieve an average accuracy of 96.59%. This helps our system perform well for disabilities to grasp objects of different sizes and shapes adaptively. Finally, a test of the people with disabilities grasping 15 objects of different sizes and shapes was carried out and achieved good experimental results.

Adaptive Neural Network Control of a Flexible Spacecraft Subject to Input Nonlinearity and Asymmetric Output Constraint.

This article focuses on the vibration reducing and angle tracking problems of a flexible unmanned spacecraft system subject to input nonlinearity, asymmetric output constraint, and system parameter uncertainties. Using the backstepping technique, a boundary control scheme is designed to suppress the vibration and regulate the angle of the spacecraft. A modified asymmetric barrier Lyapunov function is utilized to ensure that the output constraint is never transgressed. Considering the system robustness, neural networks are used to handle the system parameter uncertainties and compensate for the effect of input nonlinearity. With the proposed adaptive neural network control law, the stability of the closed-loop system is proved based on the Lyapunov analysis, and numerical simulations are carried out to show the validity of the developed control scheme.

A 3-DOF Bionic Waist Joint for Humanoid Robot.

In this study, a 3 degree-of-freedom bionic waist joint was developed with coupled tendon-driven mechanism. This bionic waist joint can not only ensure the safety of the human-robot interface, but also increase the load capacity without increasing the weight. The coupled tendon-driven mechanism enables the motion of each joint to be driven by at least two motors together, and enables a maximum torque of 3 times the maximum motor output torque at each joint. The bionic waist joint has similar kinematic characteristics to a human waist, including degrees of freedom (DOF) and range of motion (ROM). The problem of coexistence of coupling and decoupling in the same rotating joint was solved with a novel mechanism that can promote further versatility of the coupled tendon-driven mechanism. The basic movements and characteristics of the waist was validated in the experiment.

Investigation of Motor Point Shift and Contraction Force of Triceps Brachii for Functional Electrical Stimulation.

Functional electrical stimulation (FES) has been used for neurorehabilitation of individuals with paralysis due to spinal cord injuries or stroke aftereffects. The biceps brachii is often adopted in studies on FES because of the ease of stimulation, while there are few studies on the triceps brachii. Stimulation of the triceps brachii is important because the biceps brachii tends to be spastic. The aim of this study is to investigate the position shift of the motor points (MPs) of the three main muscle groups in triceps brachii with respect to the elbow joint angle, and the contraction force of the muscle groups. Firstly, MPs were measured in 6 healthy individuals using an MP pen at 5 elbow joint angles. The MPs of the long and lateral heads shifted distally and laterally, and the MPs of the medial head shifted distally and medially as the arm extended. The MPs of the long head shifted farthest of all. Secondly, the contraction force was measured in 9 healthy individuals using a force gauge at elbow joint angle of 90 degrees. Three different voltages were applied: 4, 8, and 12 V. The results showed that the medial head yields a sufficient contraction force although the medial head is situated deeper than the other two muscle groups. These findings will help to better understand the stimulation of the triceps brachii and improve the efficiency of electrical stimulation therapy.

A Study on the Classification Effect of sEMG Signals in Different Vibration Environments Based on the LDA Algorithm.

Myoelectric prosthesis has become an important aid to disabled people. Although it can help people to recover to a nearly normal life, whether they can adapt to severe working conditions is a subject that is yet to be studied. Generally speaking, the working environment is dominated by vibration. This paper takes the gripping action as its research object, and focuses on the identification of grasping intentions under different vibration frequencies in different working conditions. In this way, the possibility of the disabled people who wear myoelectric prosthesis to work in various vibration environment is studied. In this paper, an experimental test platform capable of simulating 0-50 Hz vibration was established, and the Surface Electromyography (sEMG) signals of the human arm in the open and grasping states were obtained through the MP160 physiological record analysis system. Considering the reliability of human intention recognition and the rapidity of algorithm processing, six different time-domain features and the Linear Discriminant Analysis (LDA) classifier were selected as the sEMG signal feature extraction and recognition algorithms in this paper. When two kinds of features, Zero Crossing (ZC) and Root Mean Square (RMS), were used as input, the accuracy of LDA algorithm can reach 96.9%. When three features, RMS, Minimum Value (MIN), and Variance (VAR), were used as inputs, the accuracy of the LDA algorithm can reach 98.0%. When the six features were used as inputs, the accuracy of the LDA algorithm reached 98.4%. In the analysis of different vibration frequencies, it was found that when the vibration frequency reached 20 Hz, the average accuracy of the LDA algorithm in recognizing actions was low, while at 0 Hz, 40 Hz and 50 Hz, the average accuracy was relatively high. This is of great significance in guiding disabled people to work in a vibration environment in the future.

Asymmetric shape of distal phalanx of human finger improves precision grasping.

In morphology field, the functions of an asymmetric-shaped distal phalanx in human finger have only been inferred. In this study, we used an engineering approach to empirically examine the effects of the shape of distal phalanx on the ability of precision grasping. Hence, we developed artificial fingertips consisting of four parts, namely bones, nails, skin, and subcutaneous tissue, that substitute the actual human fingertips. Furthermore, we proposed a method to evaluate the grasping ability of artificial fingers. When a cylindrical object was grasped by an artificial fingertip, a pull-out experiment was conducted. Thus, the asymmetric type was found to be superior in terms of drawing force, holding time, and work of friction than the symmetric type. Our results clearly demonstrate that the asymmetric shape, particularly the mirror-reversed shape of the distal phalanx, improves the ability of precision grasping and suggests that the human distal phalanx is shaped favorably for object grasping.

Joint angle based motor point tracking stimulation for surface FES: A Study on biceps brachii.

Functional electrical stimulation (FES) has been an effective treatment option in clinical rehabilitation such as motor function recovery after stroke. The main limitation of FES is the lack of stimulation efficiency in motor unit recruitment compared with voluntary contractions, which may cause the early onset of muscle fatigue. The stimulation efficiency of FES can be improved by optimizing electrode positions to target the motor point (MP). However, the location of MP relative to the skin may shift with the change of muscle geometry during dynamic exercise. Hence, the purpose of this study is to maintain the stimulation efficiency of FES in dynamic exercise by switching the stimulation position to follow the shift of MP. We first measured the shift of the MP of the biceps brachii with respect to the elbow joint angle, and then conducted an experiment to compare four stimulation methods: 2-channel simultaneous stimulation (SS), 2-channel time based shifting stimulation (TSS), 2-channel joint angle based shifting stimulation (JASS), and 3-channel JASS. TSS and JASS were designed as two different MP tracking strategies. The experimental results show that the 3-channel JASS caused the smallest decrease in the maximal elbow angle and the angular velocity. The results also suggest that MP tracking stimulation based on joint angle is effective for the sustainable induction of muscle contraction. Both tracking selectivity and tracking density were shown to be important to improve the stimulation efficiency of FES.

Survey on Main Drive Methods Used in Humanoid Robotic Upper Limbs.

Humanoid robotic upper limbs including the robotic hand and robotic arm are widely studied as the important parts of a humanoid robot. A robotic upper limb with light weight and high output can perform more tasks. The drive system is one of the main factors affecting the weight and output of the robotic upper limb, and therefore, the main purpose of this study is to compare and analyze the effects of the different drive methods on the overall structure. In this paper, we first introduce the advantages and disadvantages of the main drive methods such as tendon, gear, link, fluid (hydraulic and pneumatic), belt, chain, and screw drives. The design of the drive system is an essential factor to allow the humanoid robotic upper limb to exhibit the structural features and functions of the human upper limb. Therefore, the specific applications of each drive method on the humanoid robotic limbs are illustrated and briefly analyzed. Meanwhile, we compared the differences in the weight and payload (or grasping force) of the robotic hands and robotic arms with different drive methods. The results showed that the tendon drive system is easier to achieve light weight due to its simple structure, while the gear drive system can achieve a larger torque ratio, which results in a larger output torque. Further, the weight of the actuator accounts for a larger proportion of the total weight, and a reasonable external placement of the actuator is also beneficial to achieve light weight.

Learning Data Correction for Myoelectric Hand Based on "Survival of the Fittest".

In recent years, myoelectric hands have become multi-degree-of-freedom (DOF) devices, which are controlled via machine learning methods. However, currently, learning data for myoelectric hands are gathered manually and thus tend to be of low quality. Moreover, in the case of infants, gathering accurate learning data is nearly impossible because of the difficulty of communicating with them. Therefore, a method that automatically corrects errors in the learning data is necessary. Myoelectric hands are wearable robots and thus have volumetric and weight constraints that make it infeasible to store large amounts of data or apply complex processing methods. Compared with general machine learning methods such as image processing, those for myoelectric hands have limitations on the data storage, although the amount of data to be processed is quite large. If we can use this huge amount of processing data to correct the learning data without storing the processing data, the machine learning performance is expected to improve. We then propose a method for correcting the learning data through utilisation of the signals acquired during the use of the myoelectric hand. The proposed method is inspired by "survival of the fittest." The effectiveness of the method was verified through offline analysis. The method reduced the amount of learning data and learning time by approximately a factor of 10 while maintaining classification rates. The classification rates improved for one participant but slightly deteriorated on average among all participants. To solve this problem, verifying the method via interactive learning will be necessary in the future.

Semi-Automated Control System for Reaching Movements in EMG Shoulder Disarticulation Prosthesis Based on Mixed Reality Device.

Goal: The development of a control system for an electromyographic shoulder disarticulation (EMG-SD) prosthesis to rapidly achieve a task with a reduction in the operational failure of the user. Methods: The motion planning of an EMG-SD prosthesis was automated using measured visual information through a mixed reality device. The detection of an object to be grasped and motion execution depended on the EMG of the user, which gives voluntary controllability and makes the system semi-automated. Two evaluation experiments with reaching and reach-to-grasp movements were conducted to compare the performance of the conventional system when operated using only visual feedback control of the user. Results: The proposed system can more rapidly and accurately achieve reaching movements (32% faster) and more accurate (69%) reach-to-grasp movements than a conventional system. Conclusions: The proposed control system achieves a high task performance with a reduction in the operational failure of an EMG-SD prosthesis user.

Development of a Shoulder Disarticulation Prosthesis System Intuitively Controlled With the Trunk Surface Electromyogram.

We developed an intuitively operational shoulder disarticulation prosthesis system that can be used without long-term training. The developed system consisted of four degrees of freedom joints, as well as a user adapting control system based on a machine learning technique and surface electromyogram (EMG) of the trunk. We measured the surface EMG of the trunk of healthy subjects at multiple points and analyzed through principal component analysis to identify the proper EMG measurement portion of the trunk, which was determined to be distributed in the chest and back. Additionally, evaluation experiments demonstrated the capability of four healthy subjects to grasp and move objects in the horizontal as well as the vertical directions, using our developed system controlled via the EMG of the chest and back. Moreover, we also quantitatively confirmed the ability of a bilateral shoulder disarticulation amputee to complete the evaluation experiment similar to healthy subjects.