Comparison of six electromyography acquisition setups on hand movement classification tasks

Pizzolato, Stefano; Tagliapietra, Luca; Cognolato, Matteo; Reggiani, Monica; Müller, Henning; Atzori, Manfredo

doi:10.1371/journal.pone.0186132

Cited by 308 publications

(277 citation statements)

References 41 publications

Supporting

Mentioning

253

Contrasting

Order By: Relevance

“…The implementation employed for all the classifiers comes from the scikit-learn (v.1.13.1) Python package [43]. The four feature sets employed for comparison purposes are: 1) Time Domain Features (TD) [37]: This set of features, which is probably the most commonly employed in the literature [29], often serves as the basis for bigger feature sets [1], [39], [34]. As such, TD is particularly well suited to serve as a baseline comparison for new classification techniques.…”

Section: A Feature Setsmentioning

confidence: 99%

“…The feature instead calculates the cumulative energy of each level of the decomposition. The computation of the mDWT for each channel is implemented as follow in [34] (See Algorithm 1). M xk.append(val) return M xk Where x is the 1-d signal from which to calculate the mDWT and wavDec is a function that calculates the wavelet decomposition of a vector at level n using the wavelet wav.…”

Section: B Frequency Domain Featuresmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning for Electromyographic Hand Gesture Signal Classification Using Transfer Learning

Côté-Allard¹,

Fall²,

Drouin³

et al. 2019

IEEE Trans. Neural Syst. Rehabil. Eng.

567

378

View full text Add to dashboard Cite

In recent years, deep learning algorithms have become increasingly more prominent for their unparalleled ability to automatically learn discriminant features from large amounts of data. However, within the field of electromyographybased gesture recognition, deep learning algorithms are seldom employed as they require an unreasonable amount of effort from a single person, to generate tens of thousands of examples.This work's hypothesis is that general, informative features can be learned from the large amounts of data generated by aggregating the signals of multiple users, thus reducing the recording burden while enhancing gesture recognition. Consequently, this paper proposes applying transfer learning on aggregated data from multiple users, while leveraging the capacity of deep learning algorithms to learn discriminant features from large datasets. Two datasets comprised of 19 and 17 able-bodied participants respectively (the first one is employed for pre-training) were recorded for this work, using the Myo Armband. A third Myo Armband dataset was taken from the NinaPro database and is comprised of 10 able-bodied participants. Three different deep learning networks employing three different modalities as input (raw EMG, Spectrograms and Continuous Wavelet Transform (CWT)) are tested on the second and third dataset. The proposed transfer learning scheme is shown to systematically and significantly enhance the performance for all three networks on the two datasets, achieving an offline accuracy of 98.31% for 7 gestures over 17 participants for the CWT-based ConvNet and 68.98% for 18 gestures over 10 participants for the raw EMG-based ConvNet. Finally, a use-case study employing eight able-bodied participants suggests that real-time feedback allows users to adapt their muscle activation strategy which reduces the degradation in accuracy normally experienced over time.

show abstract

Section: A Feature Setsmentioning

confidence: 99%

Section: B Frequency Domain Featuresmentioning

confidence: 99%

Deep Learning for Electromyographic Hand Gesture Signal Classification Using Transfer Learning

Côté-Allard¹,

Fall²,

Drouin³

et al. 2019

IEEE Trans. Neural Syst. Rehabil. Eng.

567

378

View full text Add to dashboard Cite

show abstract

“…In this work, the fourth version of Non-Invasive Adaptive Prosthetic (NinaPro) project is employed [12]. NinaPro project is a publicity access EMG database, which records a huge number of EMG data from multiple subjects.…”

Section: A Emg Datamentioning

confidence: 99%

“…In the experiment, each motion was www.ijacsa.thesai.org performed for 5 seconds, followed by a resting state of 3 seconds. Additionally, each motion was repeated for six times, and the EMG signals were sampled at 2 kHz [12]. Furthermore, all the resting phases are removed.…”

Section: A Emg Datamentioning

confidence: 99%

A Comparative Analysis of Wavelet Families for the Classification of Finger Motions

Too¹,

Abdullah²,

Saad³

2019

IJACSA

View full text Add to dashboard Cite

Wavelet transform (WT) has been widely used in biomedical, rehabilitation and engineering applications. Due to the natural characteristic of WT, its performance is mostly depending on the selection of mother wavelet function. A proper mother wavelet ensures the optimum performance; however, the selection of mother wavelet is mostly empirical and varies according to dataset. Hence, this paper aims to investigate the best mother wavelet of discrete wavelet transform (DWT) and wavelet packet transform (WPT) in the classification of different finger motions. In this study, twelve mother wavelets are evaluated for both DWT and WPT. The electromyography (EMG) data of 12 finger motions are acquired from online database. Four useful features are extracted from each recorded EMG signal via DWT and WPT transformation. Afterward, support vector machine (SVM) and linear discriminate analysis (LDA) are employed for performance evaluation. Our experimental results demonstrate Bior3.3 to be the most suitable mother wavelet in DWT. On the other hand, WPT with Bior2.2 overtakes other mother wavelets in the classification of finger motions. The results obtained suggest that Biorthogonal families are more suitable for accurate EMG signals classification.

show abstract

“…A few years ago, a lowcost gesture recognition armband called Myo was released by Thalmic Labs (Ontario, Canada) 5 , costing approximately 200 dollars instead of other control systems that costs in the range of $10,000. The Myo armband was successfully applied in hand gesture recognition tasks in intact and hand amputated subjects [17], [18], [19], [20]. The performance of the Myo for gesture classification tasks was compared with 5 other commonly used acquisition setups on a standardized data acquisition and analysis protocol, showing that the classification accuracy is comparable to expensive setups in a two-armband configuration and moderately lower in the standard configuration [18].…”

Section: Introductionmentioning

confidence: 99%

Multifunction control and evaluation of a 3D printed hand prosthesis with the Myo armband by hand amputees

Cognolato

Atzori

Marchesini

et al. 2018

Preprint

View full text Add to dashboard Cite

Upper limb amputations are highly impairing injuries that can substantially limit the quality of life of a person. The most advanced dexterous prosthetic hands have remarkable mechanical features. However, in most cases, the control systems are a simple extension of basic control protocols, making the use of the prosthesis not intuitive and sometimes complex. Furthermore, the cost of dexterous prosthetic hands is often prohibitive, especially for the pediatric population and developing countries. 3D printed hand prostheses can represent an opportunity for the future. Open 3D models are increasingly being released, even for dexterous prostheses that are capable of moving each finger individually and actively rotating the thumb. However, the usage and test of such devices by hand amputees (using electromyography and classification methods) is not well explored. The aim of this article is to investigate the usage of a cost-effective system composed of a 3D printed hand prosthesis and a low-cost myoelectric armband. Two subjects with transradial amputation were asked to wear a custom-made socket supporting the HANDi Hand and the Thalmic Labs Myo armband. Afterwards, the subjects were asked to control and use the prosthetic hand to grasp several objects by attempting to perform a set of different hand gestures. Both the HANDi Hand and the Myo armband performed well during the test, which is encouraging considering that the HANDi Hand was developed as a research platform. The results are promising and show the feasibility of the multifunction control of dexterous 3D printed hand prostheses based on low-cost setups. Factors as the level of the amputation, neuromuscular fatigue and mechanical limitations of the 3D printed hand prosthesis can influence the performance of the setup. Practical aspects such as usability and robustness will need to be addressed for successful application in daily life. A video of the tests can be found at the following link: https://youtu.be/iPSCAbd17Qw

show abstract

Comparison of six electromyography acquisition setups on hand movement classification tasks

Cited by 308 publications

References 41 publications

Deep Learning for Electromyographic Hand Gesture Signal Classification Using Transfer Learning

Deep Learning for Electromyographic Hand Gesture Signal Classification Using Transfer Learning

A Comparative Analysis of Wavelet Families for the Classification of Finger Motions

Multifunction control and evaluation of a 3D printed hand prosthesis with the Myo armband by hand amputees

Contact Info

Product

Resources

About