Figure 2 - uploaded by Maria Claudia F Castro
Content may be subject to copyright.
Finger positioning: (a) Thumb, (b) Index finger, (c) Middle finger, (d) Ring finger, (e) Little finger.

Finger positioning: (a) Thumb, (b) Index finger, (c) Middle finger, (d) Ring finger, (e) Little finger.

Source publication
Article
Full-text available
Background Myoelectric controlled prosthetic hand requires machine based identification of hand gestures using surface electromyogram (sEMG) recorded from the forearm muscles. This study has observed that a sub-set of the hand gestures have to be selected for an accurate automated hand gesture recognition, and reports a method to select these gestu...

Similar publications

Article
Full-text available
Recently, surface electromyography (sEMG) emerged as a novel biometric authentication method. EMG system parameters, such as the feature sets and channel numbers, have been known to affect system performances. Therefore, it is crucial to investigate these effects on the performance of the sEMG-based biometric system to determine optimal system para...
Article
Full-text available
The interaction between humans and computers has become more necessary and more specific. Thumb as the most important finger plays a decisive role in decoding the gesture, especially in controlling smart phones and many other smart devices, etc. As a result, this study aims to decode the different thumb gestures from sEMG signal and to improve the...
Article
Full-text available
recent years, robots have been introduced in most factories. However, manual work still continues to be done in some places where giant robots cannot be installed. In particular, traditional Japanese crafts are done by hand, and people that engage in such crafts are called craftsmen. Generally, such artisans need years of training and cannot become...
Article
Full-text available
Finger movement identification is an important innovative interfacing method which has countless possible applications. It can be used to create a new age in human computer interfacing (HCI) devices. It can also be applied to medical applications, such as in the development of a more advanced prosthetic hand. The current research for this purpose i...
Chapter
Full-text available
Deep learning (DL) has transformed the field of data analysis by dramatically improving the state of the art in various classification and prediction tasks. Especially in the area of computer vision and speech processing, DL has recently demonstrated better performance and generalisation properties, compared to classical machine learning approaches...

Citations

... In a similar fashion, Li et al. [55] used five pairs of wet electrodes and benchtop electronics to control an aerial drone in four directions with four coarse but unique hand gestures (fist, wrist flex, wrist up, and ring finger flex) (Figure 3b). More complex hand and finger gestures detection have been widely demonstrated, while exploring different forearm electrode placements for optimal recognition accuracy [56]- [58]. ...
... The set of 6 gestures retained for the current study is pictured on Fig. 3a. It is specifically curated from the typically useful functional hand prosthesis grip modes 4,26 . It reflects common prosthesis user needs and provides the functionality expected in a commercial prosthesis. ...
Article
Full-text available
Myoelectric hand prostheses offer a way for upper-limb amputees to recover gesture and prehensile abilities to ease rehabilitation and daily life activities. However, studies with prosthesis users found that a lack of intuitiveness and ease-of-use in the human-machine control interface are among the main driving factors in the low user acceptance of these devices. This paper proposes a highly intuitive, responsive and reliable real-time myoelectric hand prosthesis control strategy with an emphasis on the demonstration and report of real-time evaluation metrics. The presented solution leverages surface high-density electromyography (HD-EMG) and a convolutional neural network (CNN) to adapt itself to each unique user and his/her specific voluntary muscle contraction patterns. Furthermore, a transfer learning approach is presented to drastically reduce the training time and allow for easy installation and calibration processes. The CNN-based gesture recognition system was evaluated in real-time with a group of 12 able-bodied users. A real-time test for 6 classes/grip modes resulted in mean and median positive predictive values (PPV) of 93.43% and 100%, respectively. Each gesture state is instantly accessible from any other state, with no mode switching required for increased responsiveness and natural seamless control. The system is able to output a correct prediction within less than 116 ms latency. 100% PPV has been attained in many trials and is realistically achievable consistently with user practice and/or employing a thresholded majority vote inference. Using transfer learning, these results are achievable after a sensor installation, data recording and network training/fine-tuning routine taking less than 10 min to complete, a reduction of 89.4% in the setup time of the traditional, non-transfer learning approach.
... The accuracy of the proposed system was comparable to previous multichannel channel EMG based classification. [16,17,19,45] This approach reduces the cost and complexity of the system and is more practical for hand prosthesis control. ...
Article
Patients suffering from upper limb amputation experience a significant reduction in their ability to perform various tasks. The currently available myoelectric prostheses seek to reinstate the lost abilities of the amputees. However, most users abandon their devices due to high cost, large size, heavy weight, complexity, and limited functional control. This paper presents a multifunctional prosthetic hand that can perform six hand activities deploying a single-channel surface electromyography (sEMG) sensor. EMG signals for fifteen subjects (five amputees and ten intact) were acquired for six contraction levels of forearm muscles using the designed sensor. These levels were further classified to recognize six predefined hand gestures by a fuzzy logic classifier. The proposed system showed an excellent success (> 95%) and other performance parameters above 96%. The intended classification-based control scheme was further realized in real-time to achieve six grip patterns for the developed prosthetic prototype. The hand was tested on five subjects (two amputees and three intact) showed a percentage of success above 91% for accomplishing the dexterous grasping operations. The proposed approach is modest, efficient, and provides a low-cost solution to amputees with multiple degrees of freedom with a single channel EMG system.
... Recognizing multiple gestures is a challenge as the recognition accuracy decreases with the increase in the number of gestures. Table 1 summarizes examples of previous literature attempting to recognize several gestures [10,[14][15][16][17][18][19][20][21][22][23][24][25]. ...
Article
Full-text available
Myoelectric prostheses are a viable solution for people with amputations. The challenge in implementing a usable myoelectric prosthesis lies in accurately recognizing different hand gestures. The current myoelectric devices usually implement very few hand gestures. In order to approximate a real hand functionality, a myoelectric prosthe-sis should implement a large number of hand and finger gestures. However, increasing number of gestures can lead to a decrease in recognition accuracy. In this work a Myo armband device is used to recognize fourteen gestures (five build in gestures of Myo armband in addition to nine new gestures). The data in this research is collected from three body-able subjects for a period of 7 seconds per gesture. The proposed method uses a pattern recognition technique based on Multi-Layer Perceptron Neural Network (MLPNN). The results show an average accuracy of 90.5% in recognizing the proposed fourteen gestures.
... However, due to the poor quality of the acquired sEMG signals from the remnant muscles of the amputated limb, the real-time classification becomes very challenging [9]. To overcome this, there have been several approaches [10]- [14] reported thus far which can be broadly classified into the following four categories: (i) reduce the number of movements considered for classification, (ii) increase the number of body-worn electrodes, (iii) use of implantable electrodes and (iv) use of the multi-modal approach. ...
... Surface electrodes have the advantage of being noninvasive and measure gross estimate of the muscle activity, but lack specificity. Reduction of the number of movements by the user improves the specificity of detection of the command from the signals [10]. However, this is at the cost of reduced dexterity and naturalness for the user, and thus user satisfaction. ...
... The study by [20] has shown an increase in performance compared to the proposed LoCoMo-Net model by 4.4%. However, this cannot be a fair comparison owing to the fact that they have evaluated their model on diffeent dataset, having less number of movement types compared to the propsoed methodology [10]. The parameters of the deep-learning models such as input size, input type (1D or 3D), filter size, number of filters, convolution layers, pooling layers, and the fully connected layer directly contribute to the computational complexity of the model as given in Table 1 and Fig. 6. ...
Article
Full-text available
Background: The enhancement in the performance of the myoelectric pattern recognition techniques based on deep learning algorithm possess computationally expensive and exhibit extensive memory behavior. Therefore, in this paper we report a deep learning framework named 'Low-Complex Movement recognition-Net' (LoCoMo-Net) built with convolution neural network (CNN) for recognition of wrist and finger flexion movements; grasping and functional movements; and force pattern from single channel surface electromyography (sEMG) recording. The network consists of a two-stage pipeline: 1) input data compression; 2) data-driven weight sharing. Methods: The proposed framework was validated on two different datasets- our own dataset (DS1) and publicly available NinaPro dataset (DS2) for 16 movements and 50 movements respectively. Further, we have prototyped the proposed LoCoMo-Net on Virtex-7 Xilinx field-programmable gate array (FPGA) platform and validated for 15 movements from DS1 to demonstrate its feasibility for real-time execution. Results: The effectiveness of the proposed LoCoMo-Net was verified by a comparative analysis against the benchmarked models using the same datasets wherein our proposed model outperformed Twin- Support Vector Machine (SVM) and existing CNN based model by an average classification accuracy of 8.5 % and 16.0 % respectively. In addition, hardware complexity analysis is done to reveal the advantages of the two-stage pipeline where approximately 27 %, 49 %, 50 %, 23 %, and 43 % savings achieved in lookup tables (LUT's), registers, memory, power consumption and computational time respectively. Conclusion: The clinical significance of such sEMG based accurate and low-complex movement recognition system can be favorable for the potential improvement in quality of life of an amputated persons.
... The overall percentage of success for correctly recognizing each gesture for 360 attempts was attained more than 97 %. The proposed system presented comparable accuracy to that of multi-channel EMG classification performed by researchers [39,[62][63][64]. Such a single-channel based classification approach reduces the overall cost and complexity of the system and provides a more practical solution for the application of controlling hand prosthesis (Table 8). ...
Article
Existing myoelectric prostheses can provide solutions to amputees for performing activities of daily livings. However, there are several issues with these prostheses (1) performance are affected mainly by sweat, electrode shift, motion artifact, interference, etc. (2) are expensive (3) have complex control schemes. This paper presents an affordable hand prosthesis solution for transradial amputees, which uses force myography (FMG) as a control signal in place of electromyography (EMG) signal. A novel FMG sensor was designed to detect mechanical muscle contractions from the residual forearm of amputees. Two distinct control strategies were formulated for translating the FMG signals from the sensor to the control commands for driving custom-made prosthetic hands. Prosthetic hand prototype 1 with a proportional control scheme was developed and verified on five amputees. Subjects wearing the hand were able to grasp various objects using their intent of muscular contractions. Fuzzy logic based classification scheme with an offline success rate above 97 % was established for identifying six different hand gestures from only single-channel FMG data. Further, the system, along with the threshold control, was implemented for the developed hand prototype 2 to obtain six distinct grip patterns. The real-time testing of hand on thirteen subjects (five amputees and eight intact) showed a success percentage above 95 % in performing dexterous grasping operations. As compared to the presently available hands, the hand prototype 2 displayed comparable features at such a low-cost price. However, with limited functionality, hand prototype 1 is an affordable, practical, intuitive, and faster version of the prosthetic hand.
... (12) Castro et al. used five channels of sEMG signals to recognize six finger movements and achieved an accuracy of 97%. (13) Anam and Jumaily proposed a new extreme learning machine and a new dimension reduction method called the spectral regression extreme learning machine for finger movement recognition. (14) Five to eight finger movements were recognized using two-channel sEMG signals. ...
... Devices and their software applications have more functionalities, and more complex systems are needed to control or access these functions. Currently, devices are being developed that are controlled with body movements, some using cameras and images [1,2] and others using EMG signals directly [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. These signals are used to estimate strength and changes in muscle activation caused by neuromuscular abnormalities and the control of medical devices [10]. ...
Chapter
Full-text available
Surface electromyography (sEMG) signals offer information on the natural control of muscle contraction but struggle to identify temporal pattern parameters for several degrees of motion of voluntary hand movements. The complex nature of these signals renders the movement prediction task difficult; therefore, feature extraction and selection algorithms are a natural choice to transform time domain data into a new space domain to enhance recognition. The purpose of this work was to conduct an analysis of a former forearm sEMG database to improve a model to classify 15 defined hand movements. A simpler classification model was created from algorithms, such as naive Bayes (NB), linear discriminant analysis (LDA), and quadratic discriminant analysis (QDA). Also, novel preprocessing of the EMG signal data was employed and modeled the movement in virtual simulation software. In the preprocessing, outliers were eliminated, and a scatter matrix algorithm was used to transform the data into a new space to increase the differentiation between distinct classes. The processing window was 62.5 ms to generate a classification and integrate one video frame movement. Experiments yielded promising results, achieving a 93.76% recognition rate in an independent test set. The biomechanical wrist model available in OpenSim was completed by adding the missing degrees of freedom of the fingers to simulate the movement generated from the proposed classification model. The sequence of movement was converted to a biomechanical model and constructed into a video object with the potential for real time use.
... While user specific control systems go some way to alleviate limitations surrounding the number of DOF, determining limits on the number of DOF to achieve the desired performance may also prove an important step in the design of prosthetic devices. The study by Castro et al. [24] recognized the need for a large number of DOF for the prosthetic device to be able to offer greater flexibility to the user but concluded that it is necessary to limit the number of DOF for accurate classification of sEMG. They attempted to maximize the sensitivity and specificity by generating a series of confusion matrices and determined that for the system to be both sensitive and specific, it was essential to determine the minimum number of actions and DOF. ...
... However, the control over griptransitions during activities of daily living remains a significant challenge for the majority of myoelectric prosthetic users to master. Intention detection based on user-modulation of sEMG, through implementation of real time classification algorithms, adaptive learning methods, binary classifications or pattern recognition (Englehart et al., 2001;Englehart and Hudgins, 2003;Ajiboye and Weir, 2005;Yonghong et al., 2005;Parker et al., 2006;Amsüss et al., 2014;Castro et al., 2015) can support effective object handling and manipulation in expert users of myoelectric prostheses. Yet, despite the potential functional gains these devices can provide to experienced users, control for many Flexor digitorium superficialis (FDS); (e). ...
Article
Full-text available
Hand gesture and grip formations are produced by the muscle synergies arising between extrinsic and intrinsic hand muscles and many functional hand movements involve repositioning of the thumb relative to other digits. In this study we explored whether changes in thumb posture in able-body volunteers can be identified and classified from the modulation of forearm muscle surface-electromyography (sEMG) alone without reference to activity from the intrinsic musculature. In this proof-of-concept study, our goal was to determine if there is scope to develop prosthetic hand control systems that may incorporate myoelectric thumb-position control. Healthy volunteers performed a controlled-isometric grip task with their thumb held in four different opposing-postures. Grip force during task performance was maintained at 30% maximal-voluntary-force and sEMG signals from the forearm were recorded using 2D high-density sEMG (HD-sEMG arrays). Correlations between sEMG amplitude and root-mean squared estimates with variation in thumb-position were investigated using principal-component analysis and self-organizing feature maps. Results demonstrate that forearm muscle sEMG patterns possess classifiable parameters that correlate with variations in static thumb position (accuracy of 88.25 ± 0.5% anterior; 91.25 ± 2.5% posterior musculature of the forearm sites). Of importance, this suggests that in transradial amputees, despite the loss of access to the intrinsic muscles that control thumb action, an acceptable level of control over a thumb component within myoelectric devices may be achievable. Accordingly, further work exploring the potential to provide myoelectric control over the thumb within a prosthetic hand is warranted.