Training with 50 sets of data as previously described was carried out, and ANN Backpropagation algorithm learning outcomes were tested with 15 sets of test data. In this test, the accuracy of the classification results of the ANN Backpropagation algorithm was compared for the ability in classifying provided test data correctly. In case of the given test data is a myoelectric signal of grasping movement, then the conclusion of ANN Backpropagation algorithm should be grasping. This was considered that ANN Backpropagation algorithm could classify correctly. This was applied similarly for grip and pinch movements. Experimental results can be seen in Fig. 8. Extraction feature method 1 showed an 86% accuracy for grasping, 93% for pinching and 100% for gripping. Extraction feature method 2 showed 86% accuracy for grasping, 60% for pinching, and 100% for gripping. Therefore, extraction features method 1, developed by Rillo et al., showed better accuracy, and then it was used as the extraction feature method in the implementation of prosthetic hand movement control. B. Prosthetic Hand's Real-Time Control Examination
Abstract: This paper presents a study of the classification of myoelectric signal using spectrogram with different window sizes. The electromyography (EMG) signals of 40 hand movement types are collected from 10 subjects through NinaPro database. By employing spectrogram, the EMG signals are represented in time-frequency representation. Ten features are extracted from spectrogram for performance evaluation. In this study, two classifiers namely support vector machine (SVM) and linear discriminate analysis (LDA) are used to evaluate the performance of spectrogram features in the classification of EMG signals. To determine the best window size in spectrogram, three different Hanning window sizes are examined. The experimental results indicate that by applying spectrogram with optimize window size and LDA, the highest mean classification accuracy of 91.29% is obtained.
Implanted motor neuroprostheses offer significant restoration of function for individuals with spinal cord injury. Providing adequate user control for these devices is a challenge but is crucial for successful performance. Electromyographic (EMG) signals can serve as effective control sources, but the number of above-injury muscles suitable to provide EMG-based control signals is very limited. Previous work has shown the presence of below-injury volitional myoelectric signals even in subjects diagnosed with motor complete spinal cord injury. In this case report, we present a demonstration of a hand grasp neuroprosthesis being controlled by a user with a C6 level, motor complete injury through EMG signals from their toe flexor. These signals were successfully translated into a functional grasp output, which performed similarly to the
The classification of each movement occurred during the data acquisition. The signal was acquired during the following simple movements: contraction of the hand, wrist extension, wrist flexion, forearm flexion and fore- arm rotation. Was also carried out an assay for the acqui- sition of a complex movement: forearm rotation along with the movement of contraction of the hand. For train- ing the neural network was used one session for each move with 5 repetitions, and the training time lasted less than 1 second. The training uses the Levemberg-Marquardt algorithm and for the calculation of performance was used the technique of the mean squared error. The train- ing was made for the two subjects separately. The subject one reached the mean squared error of 0.0165 as shown in Figure 7 and the subject 2 achieved an error of 0.0531.
myoelectric signal of muscle biceps brachii from their unaffected side. Stroke survivors can use their unaffected arm to control their affected side for theraphy and this strategy could give a significant improvement in their muscle –. Therefore, processing the myoelectric signal that separated from noise generated during EMG recording became important in developing the rehabilitation devices based on myoelectric control. In this prototypye, some filters were added in signal processing to reduce noise, such as the IIR (Infinite Impulse Response) low pass filters and Kalman filter for smoothing the signals at low frequency thus provide better control to controll the movement of the post- stroke rehabilitation device. It is expected that this device can help stroke survivors to perform therapy independently without depending on therapists thus rehabilitation will be more effective and efficient.
Traditionally power spectra, derived from Fourier transforms, have been used to identify changes in the frequency content of the myoelectric signal, as they enable the calculation of mean and/or median frequency values. Such values provide an initial assessment of the frequency content of the signal and, in the past, have been useful indicators of muscle fatigue (Brody et al., 1991; Petrofsky, 1979) and for the identification of when different types of motor unit are active (Elert et al., 1992; Gerdle et al., 1988). In this study we applied principal component analysis to identify and quantify differences in myoelectric signal intensity and frequency content between different populations of fibre types. When the data set was partitioned into complete strides the first three principal components were able to describe over 95% of the signal, matching previous reports (von Tscharner, 2002; Wakeling and Rozitis, 2004; Wakeling et al., 2006). The results of this study and previous work (Wakeling and Rozitis, 2004) show that PCI closely matches the mean intensity spectrum of the myoelectric signal and therefore represents the intensity of the signal, while PCII represents the relative contribution of each frequency component to the signal. Each measured myoelectric intensity spectrum can be reconstructed by the linear combination of the PC weightings and the PC loading scores. The relative contribution of the PCI and PCII loading scores would lead to a skewing of the myoelectric spectra to lower or higher components (Eqn·3). This is demonstrated by the frequency content of the soleus and medial gastrocnemius muscle, which have, respectively, the lowest and highest mean frequency values and PCII loading scores (Fig.·5). The third principal component had two negative phases with a positive phase occurring in the mid-range of frequencies (200–400·Hz). The
13 Read more
Electrode placement is important when using any of the electrode types described above. With the bipolar electrode the optimal position of the electrodes is parallel to the muscle fibers in order to maximize the probability of reading the same signal. Here to mimic the natural hand movement myoelectric signals are collected from wrist flexor and extensor since these muscle groups are directly responsible for the palm and wrist movements of interest. Four differential myoelectric signal channels were recorded using surface electrodes. Two channels were used to record potentials from the flexors and the other two are used to record the triceps activity. The desired position for electrodes is on the belly of the muscle and not on the outer edge of the muscle where other muscles could interfere with the muscle under examination.
10 Read more
In this and a companion paper (Hodson-Tole and Wakeling, 2008) we aim to: (1) describe changes in motor unit recruitment patterns in response to changes in locomotor velocity and incline; (2) determine what underlying factors may influence the recorded motor unit recruitment patterns (Hodson-Tole and Wakeling, 2008). In this first study we therefore aim to quantify patterns of motor unit recruitment within the three ankle extensor muscles of the rat (Rattus norvegicus) during treadmill locomotion at different velocities and inclines. The rat was chosen as a suitable model for these studies because it is easily trained to perform a range of tasks and it is possible to take detailed measurements of the rat system using a combination of sonomicrometry and electromyography (Gillis and Biewener, 2001). There is also a wide range of information available on the contractile properties of rat muscles (Close, 1964; Close and Luff, 1974; Schiaffino and Reggiani, 1996). The soleus, plantaris and medial gastrocnemius muscles were studied as these contain predominantly slow, fast and a mixed population of muscle fibre types, respectively (Armstrong and Phelps, 1984). Comparing synergistic muscles with such differences in fibre type proportions, and hence mechanical properties, facilitated analysis of recruitment strategies within distinctly different fibre populations during different locomotor tasks. It was hypothesised that increasing locomotor velocity and incline would lead to an increase in myoelectric intensity in each muscle. In addition, it was hypothesised that increased locomotor velocity would lead to an increase in the high frequency myoelectric signal component from each muscle. An increase in locomotor incline was hypothesised to lead to an increase in the low frequency component of the myoelectric signal from each muscle, as we predicted the recruitment of slower motor units to provide the additional forces required for the mechanical work of overcoming gravity.
11 Read more
Intuitive myoelectric prosthesis control is difcult to achieve due to the absence of proprioceptive feedback, which forces the user to monitor grip pressure by visual information. Existing myo- electric hand prostheses form a single degree of freedom pincer motion that inhibits the stable prehension of a range of objects. Multi-axis hands may address this lack of functionality, but as with multifunction devices in general, serve to increase the cognitive burden on the user. Intelligent hierarchical control of multiple degree-of-freedom hand prostheses has been used to reduce the need for visual feedback by automating the grasping process. This paper presents a hybrid controller that has been developed to enable different prehensile functions to be initiated directly from the user’s myoelectric signal. A digital signal processor ( DSP ) regulates the grip pressure of a new six-degree- of-freedom hand prosthesis thereby ensuring secure prehension without continuous visual feedback.
generality by compensating between the user’s differences of the myoelectric signals. Therefore, it can be expected to construct the classiﬁer by using a smaller number of user’s learning data than those the traditional method needs. And the constructed classiﬁers by ANNs can be relearned using new learning data and be adjusted for the user through the relearned process. Furthermore, we construct the classiﬁer which uses Support Vector Machine (SVM) in additional to the above two methods using ANNs. We perform some experiments in order to show the validity of our considered methods. We evaluate the constructed classiﬁers in point of accuracy rate, precision rate and recall rate, and discuss our ﬁndings.
A good prosthesis design has to take into account all the problems related with the interaction between human and machines. Since the very ﬁrst need of an amputee is the social and psicological rehabilitation, patients should have a good feeling with their prosthesis. They should be able to perform daily activities without stress and ex- cessive mental load. Patient refusal, in fact, is certainly the main cause limiting the use of an active prosthesis. It depends on excessive weight, limited speed, noise, poor reliability and very high power consumption . It means that patient has to carry big batteries and can’t use the prosthesis for a long time. Otherwise, myoelectric pros- thesis are very appreciated for their easiness in controlling the movements and for the absence of wires and braces: they exploit electromyographical signals of two residual antagonist muscles of the stump to command the system. Actually, only a few externally powered elbow prostheses are commercially available: the NY Electric Elbow, the Boston Elbow, the Utah Arm, the Otto Bock Dynamic Arm  and the INAIL elbow . This work is related
Protein secondary structures are fairly well defined due to the rigid nature of the peptide bond and the free rotation about bonds either side of it. A consequence of this is that the CD spectra can be expressed as a weighted sum of the spectra corresponding to different structural motifs. Deconvo- lution of spectra can, at least in principle, be used to deter- mine the different proportions of secondary structure motifs present in the protein. For example, an a-helix is character- ized by a large positive band at 190 nm (part of the p!p * exciton couplet), and two smaller negative bands at 208 nm (the other p!p*component) and 222 nm (n !p*). b-sheets give different signals from a -helices and vary from protein to protein presumably dependent on orientation (parallel/anti- parallel), the relative size of the sheet, its three-dimensional twist. There are, however, approximate b -sheet signatures: a positive peak between 195 and 202 nm, and a negative sig- nal between 215 and 220 nm. b-turns have a large negative band at 180–190 nm, a positive signal in the 200–205 nm range ( p!p *), and a negative signal at 225 nm (n !p *). The structure often referred to as “random coil” has a negative signal at 200 nm, which is very similar to both the spectrum of a class of b -sheet proteins and poly-proline II spectra.  It
211 Read more
This paper introduces a family of multivariate distributions based on Edgeworth and Gram-Charlier expansions. This family encompasses most of the univariate densities proposed in financial literature (e.g. the so-called Gram-Charlier, Edgeworth-Sargan or Positive Edgeworth-Sargan), which can be obtained as the marginal densities of the different densities nested in this family. Therefore, the MGC densities inherit the properties of their univariate precursories in terms of flexible parameter structure to accurately represent all the characteristic features of most high-frequency financial variables (i.e. thick tails, sharp peak, asymmetries, conditional heteroskedasticity, etc.). Within this family, the specifications that are positive for all the values of the parametric space (and are thus properly defined) merit particular interest. We provide some examples of positive multivariate densities overcoming the deficiencies of the MES density, which can be understood as applications of the Gallant and Nychka (1987) methodology to the multivariate framework. The performance of these densities is compared to fit and forecast the full density of a portfolio of asset returns, and it is found that they perform quite satisfactorily and are superior to the MN, the most commonly used distribution in financial risk management.
33 Read more
MES model was used for the evaluation of the anticonvulsant effect of ALEE. Electro Convulsiometer (Model No EC-02) was used for delivering an electric shock (50 mA for 0.2 seconds) with the help of corneal electrode to induce hind limb tonic extension (HLTE) in mice (Kulkarni, 1999; Swinyard et al., 1952). ALEE was administered at the dose of 200, 400, and 600 mg/kg, orally while phenytoin (25 mg/kg, intraperitoneally) was used as a standard drug. All the treatments were given 30 minutes before applying electric shock. Animals were divided into five groups, each group containing 10 mice.
In this study, a comprehensive field study was conducted to make a preliminary hazard assessment on the Mes¸elik cam- pus area, Eskis¸ehir, Turkey. In this context, the experimen- tal studies were performed in two stages. In the first stage, boreholes were drilled in the field; a standard penetration test (SPT) was performed and disturbed/undisturbed samples were collected from certain levels. In the second stage, labo- ratory tests were performed in order to identify and classify the samples. Unconfined compression strength and triaxial compression tests were conducted on undisturbed samples for determining the engineering characteristics. XRD (X-ray diffraction) tests were performed and the swelling potential of the samples were evaluated. The liquefaction potential of the area was also assessed on a SPT-based method. Thus, the geotechnical parameters and the liquefaction potential of the sub-surface in the study area were thoroughly analyzed and presented to be used for further studies.
10 Read more
Recently, a study by Bouwsema was the first to report on visuomotor behaviours in upper limb prosthesis users . This study quantified the level of skill in myoelectric prosthesis users through exploring the relationship be- tween the clinical outcomes and different visuomotor indices. In this study, six experienced trans-radial amputees were required to perform reach to grasp and manipulation tasks with four objects (each object consisted of 2 identical- sized metal plates, separated by springs of differing stiffness. Participants were required to perform each grasp of an object using either a direct, or indirect approach. During each task, performance was evaluated based on analysis of gaze behaviour, joint angle, aperture trajectories and object compression force during manipulation. For com- parison purposes, subjects also performed the Southampton Hand Assessment Procedure (SHAP) . The study charac- terised gaze behaviour using a simple coding scheme in which the scene, recorded by a head-mounted camera, was divided into a number of categorical areas (hand, object, object and hand, endpoint and other). The authors reported time spent focusing on each of the areas in the scene and number of fixations per trial. The authors reported that all subjects focused gaze on the object being grasped for the majority of the task time, irre- spective of their performance on the SHAP test. Two subjects also tended to flick back and forth between the object and the hand during task performance. This study was the first to show that the gaze patterns of users of myoelectric prostheses differ markedly from those seen in anatomically intact subjects.
12 Read more
This analysis makes use of a bias-free method, commonly known as “blind” analysis. In a blind analysis the physics result, i.e. the signal region, is hidden until the fit model is validated. The major advantage of a blind analysis is to minimise the potential for experimenter’s bias in the result . In a blind analysis, the selection and fitting model are constructed and tested using Monte Carlo (MC) simulation to simulate the signal region and background contributions from other BB decays. Data taken below the Υ (4S) resonance, also known as “offpeak” data, is used to estimate the continuum background, i.e. background from other qq events. The MC simulation process is split in two stages: the event generation and the detector response to the passage of particles through the sub-detectors, with subsequent particle decays and electronic response.
201 Read more
The study was approved by the University of Salford Research Ethics committee (Ref # REPN09/174) and NHS National Research Ethics Service (Ref # 11/NW/0060). Seven anatomically intact individuals (four males and three females; age mean ±1standard deviation (SD): 36 ± 10 years; age range: 26-48 years) and four users of myoelectric prostheses (3 males and 1 female; age mean ±1 SD: 49 ± 10 years; age range: 35-56 years; years since myoelectric prosthesis prescription: mean ±1 SD: 20 ± 13 years, range: 2-32 years) agreed to participate in the study and gave informed consent. Of the anatomically intact individuals, six subjects were right handed and one subject was left handed. All four of the myoelectric prosthesis users were right side affected, and for three of them (S1, S2, S4) the prosthesis replaced their original dominant hand. Three subjects (S1-S3) used an Otto Bock Sensor Hand Speed and S4 used an RSL Steeper Multi- Control Plus hand. S2 and S4 were fitted with a powered wrist rotator. All subjects used a two-site two-state control strategy. All subjects were able to complete upper limb functional tasks comfortably without glasses or contact lenses. All data were collected in the Movement Science Laboratory at the University of Salford, Salford, Greater
11 Read more
Previous studies have shown that microemboli to brain occurs following an ischemic stroke [1-12]. These clini- cally silent microemboli can be detected as microem- bolic signals (MES) using transcranial Doppler (TCD). There is a wide variation in the prevalence of MES after stroke. A pooled analysis of ischemic stroke patients with a known source of embolism have shown that the prevalence of MES in symptomatic ICA stenosis, asymp- tomatic ICA stenosis and aortic atheroma as 42%, 8% and 32% respectively . However, the prevalence of MES also depends on timing of monitoring, showing higher prevalence when microemboli-monitoring is done closer to stroke onset [2,3,10]. But studies to assess the true prevalence of MES immediately following ischemic events are lacking. Also, the implications of
III. EMG E LECTRODES A ND I TS P LACEMENT EMG electrodes are used to detect the bioelectrical activates inside the muscles of the human body. The surface EMG electrodes provide a non-invasive technique for the detection and measurement of EMG signal. The EMG electrodes form a chemical equilibrium between the skin and the detecting surface by means of electrolytic conduction, so that current can flow into the electrode. Moreover Surface EMG electrodes are widely used to detect the muscle activity in order to control device to achieve prosthesis for physically disabled and amputated population. Proper skin preparation is required for the application of surface EMG electrodes. Skin impedance must be considerably reduced in order to obtain a good quality EMG signal. The dead cells on the skin e.g. hair must be completely removed from the location where the EMG electrodes are to be placed. The moisture level of the skin must be very low. Wetness or sweat on the skin can be eliminated by treating with alcohol.