Electromyographic signal integrated robot hand control for massage therapy applications

In the design of a health monitoring system, Electromyography (EMG) signal is one of the key parameters. So it is very important to utilise the EMG signal carefully. In this paper, different classification methods have been used to classify the EMG signals. EMG signals have been extracted from five different subjects corresponding to their eight different motions of the right hand using LabVIEW. The classification techniques used includes k-NN, naive Bayes and Artificial Neural Network (ANN) classifiers. Five feature vectors used are mean absolute value, average band power, standard deviation, peak to peak root mean square value and root mean square value to learn the classifier. From the results obtained, it has been observed that the performance of ANN classifier in terms of classification accuracy and time required to classify is the best among the three classifiers considered for EMG signal analysis. ANN has 100% classification efficiency for classification of EMG signals obtained from different subjects relative to their hand motion. Based upon better classification efficiency, a better health monitoring system can be manufactured.

Electromyography (EMG) signal is an myoelectric signal in the muscle layer. It occurs caused by contraction and relaxation muscle activity. This article provide numerical study of the classifying the electromyography signal for wrist movement combined with open and grasping finger flexor. The EMG signal has recorded using a device called electromyography. It has acquired by attaching an surface electrode in the skin then the electrode was capturing the raw signal. The volunteer involved were six where each volunteer has ten datasets the EMG signal. The surface electrode are sticked in the lower arm muscle. The EMG raw signal was processed using zero-mean normalization. The feature extraction method is root mean square (rms), mean absolute value (mav), variance (var), and standard deviation (std). This EMG signal has been classified by naïve bayes classifier. Training and testing data was using 5-cross validation. The result indicates that the classification accuracy for classifying the EMG signal for wrist movement combined open finger flexor (OFF) and grasping finger flexor (GFF) is 70% and 75% respectively. Therefore, the EMG signal can be applied for identificating of muscle disorder, prostheses hand and biometric system.

The purposes of this study were firstly to compare and investigate localised fatigue in the upper trapezius muscle in various arm positions as assessed by mechanomyographic (MMG) and surface electromyographic (EMG) signals and secondly to study the effects of different normalisation methods on MMG and EMG signals during non-fatiguing and fatiguing low level isometric contractions. The MMG, EMG and rate of perceived exertion were recorded from 11 subjects in five arm positions (0 degrees abduction and 0 degrees flexion, 45 degrees and 90 degrees flexion, 45 degrees and 90 degrees abduction) with different bilateral arm loads during 3 s for non-fatiguing (0-0.5-1 kg hand-load) and 3 min for fatiguing contractions (1 kg hand-load). The root mean square (RMS), average rectified value (ARV), mean power frequency (MNF), and median power frequency (MDF) of the MMG and EMG signals were computed and normalised with respect to the initial values obtained in the current arm position or in the reference position (0 degrees abduction and 0 degrees flexion) corresponding to the normal postural activity of the trapezius muscle. For fatiguing contractions, differences in magnitude of the increase in the RMS or ARV and decrease in the MNF or MDF were observed for EMG and MMG. The MMG amplitude and spectral changes followed the subjective sensation of fatigue and were not correlated to their EMG counterparts, suggesting that they may reflect different phenomena. For non-fatiguing contractions, normalisation to the current arm position entailed the loss of dynamic amplitude changes suggesting that a single reference contraction in the middle part of the range of movement is enough for proper normalisation of EMG and MMG signals. For fatiguing contractions, normalisation of the EMG and MMG to some extent can lead to a misleading interpretation. Assessment of the upper trapezius muscle by means of MMG may be valuable in ergonomics.

There is an increasing demand for accurate prediction of joint moments using wearable sensors for robotic exoskeletons to achieve precise control and for rehabilitation care to remotely monitor users’ condition. In this study, we used electromyography (EMG) signals to first identify muscle synergies, then used them to train of a long short-term memory network to predict knee joint moments during walking. Kinematics, ground reaction forces and EMG from 10 muscles on the right limb were collected from 6 able-bodied subjects during normal gait. Between 4 and 6 muscle synergies were extracted from the EMG signals, generating two outputs - the muscle synergies weight matrix and the time-dependent muscle synergies action signals. The muscle synergies action signals and measured knee joint moments from inverse dynamics were then used as inputs to train the joint moment prediction model using a long short-term memory network. For testing, between 4 and 7 EMG signals were used to estimate the muscle synergies action signals with the extracted muscle synergies weights matrix. The estimated muscle synergies action signals were then used to predict knee joint moments. Knee joint moments were also predicted directly from all 10 EMGs, then from 4-7 EMG signals using another long short-term memory network. Prediction accuracy from the synergies-trained network vs. the EMG-trained network were compared, using the same number of EMG signals in each. Prediction error with respect to moments measured via inverse dynamics was computed for both networks. Knee moments predicted with as few as 4 EMGs was at least as accurate as moments predicted from all 10 EMGs when muscle synergies were exploited. Predicted knee moments from muscle synergies achieved an average of 4.63% root mean square error from 4 EMG signals, which was lower than error when predicted directly from 4 EMG signals (5.63%).

In this paper an approach for a human robot interface (HRI) is proposed, based on electromyographic (EMG) signals interpretation, utilizing a rule-based expert system. The developed approach uses the EMG signals during the motion of the elbow and wrist joint of a human for moving the arm on a plane. After processing, these signals are passed through the rule-based expert system in order to move a KUKA LWR robot according to the movement of the human forearm. Signals from the bicep, triceps, flexor carpi, and extensor carpi muscles are extracted using four surface EMG electrodes, one in each muscle. These signals are then normalized, rectified and passed through a root mean square (RMS) algorithm twice. The main advantage of the proposed method compared to other EMG analysis and implementation is that this system makes use of only 4 EMG signals and does not need the interference of other position tracking sensors or machine learning techniques. The experimental results show that a rule-based expert system can be used adequately for the teleoperation of a two joints planar robotic arm.

This paper proposes to monitor and recognize the fatigue state during football training by analyzing the surface electromyography (EMG) signals. The surface electromyography (EMG) signal is closely connected with the state during sports and training. First, power frequency interference, motion artifacts, and baseline drift in the surface electromyography (EMG) signal are removed; second, the authors extract 6 features: rectified average value (ARV), integrated electromyography myoelectric value (IEMG), root mean square of electromyography value (RMS), median frequency (MF), average power frequency (MPF), and electromyography power (TP) to represent the surface electromyography (EMG) signal; lastly, the extracted features are input into a one-class support vector machine to determine whether the player has been fatigued and are input into a weighted support vector machine to determine the degree of fatigue if the player has been fatigued. The experimental results show that more than 95% of the fatigue state can be recognized by surface electromyography (EMG) signal.

Nowadays, analysis of electromyography (EMG) signal using wavelet transform is one of the most powerful signal processing tools. It is widely used in the EMG recognition system. In this study, we have investigated usefulness of extraction of the EMG features from multiple-level wavelet decomposition of the EMG signal. Different levels of various mother wavelets were used to obtain the useful resolution components from the EMG signal. Optimal EMG resolution component (sub-signal) was selected and then the reconstruction of the useful information signal was done. Noise and unwanted EMG parts were eliminated throughout this process. The estimated EMG signal that is an effective EMG part was extracted with the popular features, i.e. mean absolute value and root mean square, in order to improve quality of class separability. Two criteria used in the evaluation are the ratio of a Euclidean distance to a standard deviation and the scatter graph. The results show that only the EMG features extracted from reconstructed EMG signals of the first-level and the second-level detail coefficients yield the improvement of class separability in feature space. It will ensure that the result of pattern classification accuracy will be as high as possible. Optimal wavelet decomposition is obtained using the seventh order of Daubechies wavelet and the forth-level wavelet decomposition.

Relationships of Vibromyographic and Electromyographic Signals During Isometric Voluntary Contraction

Active hand prostheses are usually controlled by electromyography (EMG) signals acquired from few muscles available in the residual limb. In general, it is necessary to estimate the envelope of the EMG in real-time to implement the control of the prosthesis. Recently, sensors based on Force Sensitive Resistor (FSR) proved to be a valid alternative to monitor muscle contraction. However, FSR-based sensors measure the mechanical phenomena related to muscle contraction rather than those electrical. The aim of this study is to test the difference between the EMG and force signal in controlling a prosthetic hand. Particular emphasis has been placed on verify the prosthesis activation speed and their application to fast grabbing hand prosthesis as the "Federica" hand. Indeed, there is an intrinsic electro-mechanical delay during muscle contraction, since the electrical activation of muscle fibres always precedes their mechanical contraction. However, the EMG signal needs to be processed to control prosthesis and such filtering unavoidably causes a delay. On the contrary the force signal doesn’t need any processing. Both EMG and force signals were simultaneously recorded from the flexor carpi ulnaris muscle, while subject performed wrist flexions. The raw EMG signals were rectified and low-pass filtered to extract their envelopes. Different widespread operators were used: Moving Average, Root Mean Square, Butterworth low-pass; the cut-off frequency was set to 5 Hz. Afterward, a classic double threshold method was used to compute the muscle contraction onsets (i.e. the signal should exceed a threshold level for a certain time period). Results showed that the lag introduced by the low-pass filtering of the rectified EMG, generates delays greater than those associated with the force sensor. This analysis confirms the possibility of using force sensors as a convenient alternative to EMG signals in the control of prostheses.

The aim of this study was to investigate changes in mechanomyographic (MMG) and the surface electromyographic (EMG) signals during and after fatiguing shoulder eccentric contractions in a group consisting of 12 males and 12 females. Exerted force, MMG, EMG, pain and rate of perceived exertion were assessed before, during and after repeated high-intensity eccentric exercises. Bouts of eccentric contractions caused a decrease in the exerted force for males (P<0.05) and an increase in the rate of perceived exertion and pain for both genders (P<0.05). During eccentric exercise, the root mean square (RMS) values of the MMG signal increased (P<0.05). The mean power frequency (MPF) values of the EMG signal decreased at the end of each eccentric bout for both genders (P<0.05); the decrease was higher for females compared with males (P<0.05). Immediately after eccentric exercise in static abduction of the upper limbs, the MMG RMS and MPF values increased (P<0.05). The present study showed that (1) neuromuscular changes associated with pain and changes in muscle stiffness and (2) changes in motor units strategy during fatigue development in shoulder muscle are reflected in the MMG and EMG signals.

In this paper, the Electromyographic (EMG) signal was investigated on the Biceps Brachii muscle during dynamic contraction with two different rest intervals between trials. The EMG signal was recorded from 10 healthy right arm-dominant young subjects during load lifting task with a standard 3-kg dumbbell for 10 seconds. Root mean square (RMS) has been used to identify the muscle function. The resting period was 2- and 5-minutes between each trial. The statistical analysis techniques included in the study were i) linear regression to examine the relationship between the EMG amplitude and the endurance time, ii) repeated measures ANOVA to assess differences among the different trials and iii) the coefficient of variation (CoV) to investigate the steadiness of the EMG activation. Results show that EMG signal is more active after 5 minutes rest period compare to 2-minutes gap. On the other hand, EMG signals were steady during 2-minutes rest (7.59%) compare to 5-minutes resting interval (16.14%). Results suggest that moderate interval between each trial is better to identify the muscle activity compare to a very short interval. The findings of this study can be used to improve the current understanding of the mechanics and muscle functions of the upper limb muscle of individuals during a contraction which may prevent from muscle fatigue.

EMG is an electric signal in the human muscle layer. This signal is caused by muscle contraction activity. The main purpose of this study was to explore the pattern of electromyography signal for wrist movement in open finger extensor. The EMG can be recorded using a device called electromyography (EMG). It can be acquired by attaching an electrode to the surface of the skin and the electrode was capturing the raw of EMG signal. Volunteers involved in this study were six people where each individual have 10 datasets the EMG signals. The electrodes are installed in the lower arm muscles. The EMG raw signal was processed by normalizing zero-mean. After pre-processing, the EMG signal has been done a feature extraction process to get the EMG data which was be an input vector in Learning Vector Quantization (LVQ). The feature extraction method was mean absolute value (MAV), root mean square (RMS), minimum value (Min), maximum value (Max), variance (Var), standard deviation (STD), and length of data (LoD). This study indicates that the classification accuracy for training and testing data of the EMG signal for wrist movement in open finger extensor (OFE) and grasping finger extensor (GFE) was 70.83% and 83.33% respectively. Therefore, the EMG signal can be used for identifying muscle disorder, artificial hand control and biometric identity.

Identification of the innervation zone is widely used to optimize the accuracy and precision of noninvasive surface electromyography (EMG) signals because the EMG signal is strongly influenced by innervation zones. However, simply structured fusiform muscle, such as biceps brachii muscle, has been employed mainly due to the simplicity with which the propagation from raw EMG signals can be observed. In this study, the optimum electrode location (OEL), free from innervational influence, was investigated by the propagation pattern of action potentials for brachii muscles and more complicated deltoid muscle structures using an automatized signal analysis technique. The technique employed newly developed computer software with additional clinical uses and minimized subjective differences. EMG signals were recorded using surface array electrodes during voluntary isometric contractions obtained from 12 healthy male subjects. Peaks in EMG signals were detected and averaged for each muscle. The propagation patterns and OEL were examined from biceps brachii muscles for all subjects and from deltoid muscles for seven subjects. The estimated locations were partially confirmed by comparing the root mean squares of the EMG signals. These results show that propagation patterns and OEL could be estimated simply and automatically even from the surface EMG signals of deltoid muscles.

It may be difficult task for physically weak elderly, disabled and injured individuals to perform the day to day activities in their life. Therefore, many assistive devices have been developed in order to improve the quality of life of those people. Especially upper-limb power-assist exoskeletons have been developed since the upper limb motions are vital for the daily activities. Electromyography (EMG) signals of the upper limb muscles have sometimes been used as a primary signal to control the power assist exoskeletons since the EMG signals directly reflect the motion intention of the user. But one of the main obstacles for EMG based controller is the muscle fatigue, because the muscle fatigue might change the EMG patterns. It is important for power-assist exoskeleton to correctly assist the user for longer period of time. But it has high probability of user muscles been fatigued because users getting more and more exhausted at the end of the day. Therefore it is necessary to consider the variations of EMG signals due to the effect of muscle fatigue. In this paper it demonstrates the study which was conducted to find out the effects of muscle fatigue on the three EMG features derived from the raw EMG signals of the Bicep brachii, Deltoid-posterior, Deltoid-anterior and Supinator muscles of the upper limb. Shoulder vertical flexion/extension, shoulder abduction/adduction, elbow flexion/extension and forearm pronation/supination motions were carried out before and after a set of muscle fatiguing exercises. The three features computed in this experiment were RMS (Root Mean Square), MPF (Mean Power Frequency) and a spectral feature (FInsm5) which was proposed by Dimitrov. Comparison results of these three features of all muscles before and after the fatiguing exercises showed an percentage increase of the RMS and FInsm5 features whereas MPF showed a percentage decrease with respect to the before fatiguing conditions. The result showed that the EMG RMS may not a reliable feature to use as the only input signal in EMG based control for human upper-limb power assist in the muscle fatiguing conditions. Therefore, it is suggested that a modification method for compensating the effect of muscle fatigue is required on the EMG based control in order to have a long and reliable use of the human upper-limb power assist exoskeletons.

Control of transhumeral prosthetic devices can effectively be performed using the predicted joint torques at the elbow. The joint torque values are generally predicted using the Electromyography (EMG) signals taken from upper arm muscles of the amputee. This paper uses a Bagnoli-16 EMG system to extract EMG signals from the biceps and triceps. The EMG signals are complex to handle mainly due to the stochastic nature of the signal. Independent component analysis (ICA) is utilized to isolate the EMG signals from each muscle. In order to measure the actual torque, a novel kinematic model is proposed in this paper. For the joint torque prediction two classifiers have been developed. First an Artificial Neural Network model (ANN) based classifier is trained to predict the joint torques. Using different test data the ANN model is tested against the arm kinematic based joint torque predictions. The test results indicated 5.6% of root mean square error against the actual predicted torque values. In order to improve the classification an Artificial Neuro-Fuzzy inference system (ANFIS) has been developed. Using the same data the ANFIS based classifier produced 3.3% of the root mean square error against the kinematically predicted joint torques.