A long short-term recurrent spatial-temporal fusion for myoelectric pattern recognition

Khushaba, Rami N.; Scheme, Erik; Al-Timemy, Ali H.; Phinyomark, Angkoon; Taee, Ahmed Al; Al-Jumaily, Adel

doi:10.1016/j.eswa.2021.114977

Cited by 21 publications

(10 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results corroborate the recent literature that the traditional features, e.g. HTD and AR-RMS, can no longer compete with new features [7,13,19,23,25,45]. LSTM and CNN were significantly outperformed by their combination of CNN+LSTM on DB7 and DB5 which in turn demonstrates the power of the spatiotemporal feature learning.…”

Section: Discussionsupporting

confidence: 89%

“…The electromyographic (EMG) signal, recorded from the stump muscles, is a valuable, rich, complex, and dynamic source of information for prosthesis control. Research shows that providing natural EMG-based feedback with different approaches including grasp/wrist force and movement estimation [1][2][3][4], continuous finger trajectory decoding [5], and discrete movement classification [6,7] could achieve promising results for desirable control of the prosthesis. Machine learning is a candidate tool in mapping motor intent to prosthesis control [8][9][10].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Decoding HD-EMG Signals for Myoelectric Control - How Small Can the Analysis Window Size be?

Khushaba

Nazarpour

2021

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

Objective: The efficacy of an adopted feature extraction method directly affects the classification of the electromyographic (EMG) signals in myoelectric control applications. Most methods attempt to extract the dynamics of the multi-channel EMG signals in the time domain and on a channel-by-channel, or at best pairs of channels, basis. However, considering multi-channel information to build a similarity matrix has not been taken into account.Approach: Combining methods of long and short-term memory (LSTM) and dynamic temporal warping (DTW), we developed a new feature, called spatio-temporal warping (STW), for myoelectric signals. This method captures the spatio-temporal relationships of multi-channels EMG signals.Main Results: Across four online databases, we show that in terms of average classification error and standard deviation values, the STW feature outperforms traditional features by 5% to 17%. In comparison to the more recent deep learning models, e.g. convolutional neural networks (CNN), STW outperformed by 5% to 18%. Also, STW showed enhanced performance when compared to the CNN+LSTM model by 2% to 14%. All differences were statistically significant with a large effect size.Significance: This feasibility study provides evidence supporting the hypothesis that spatio-temporal warping of the EMG signals can enhance the classification accuracy in an explainable way when compared to recent deep learning methods. Future work includes real-time implementation of the method and testing for prosthesis control.

show abstract

Section: Discussionsupporting

confidence: 89%

Section: Introductionmentioning

confidence: 99%

Decoding HD-EMG Signals for Myoelectric Control - How Small Can the Analysis Window Size be?

Khushaba

Nazarpour

2021

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Our results corroborate the recent literature that the traditional features, e.g. HTD and AR-RMS, can no longer compete with new features [7,13,19,23,25,45].…”

Section: Discussionsupporting

confidence: 91%

Spatio-temporal warping for myoelectric control: an offline, feasibility study

2021

View full text Add to dashboard Cite

Objective. The efficacy of an adopted feature extraction method directly affects the classification of the electromyographic (EMG) signals in myoelectric control applications. Most methods attempt to extract the dynamics of the multi-channel EMG signals in the time domain and on a channel-by-channel, or at best pairs of channels, basis. However, considering multi-channel information to build a similarity matrix has not been taken into account. Approach. Combining methods of long and short-term memory (LSTM) and dynamic temporal warping, we developed a new feature, called spatio-temporal warping (STW), for myoelectric signals. This method captures the spatio-temporal relationships of multi-channels EMG signals. Main results. Across four online databases, we show that in terms of average classification error and standard deviation values, the STW feature outperforms traditional features by 5%–17%. In comparison to the more recent deep learning models, e.g. convolutional neural networks (CNNs), STW outperformed by 5%–18%. Also, STW showed enhanced performance when compared to the CNN + LSTM model by 2%–14%. All differences were statistically significant with a large effect size. Significance. This feasibility study provides evidence supporting the hypothesis that the STW feature of the EMG signals can enhance the classification accuracy in an explainable way when compared to recent deep learning methods. Future work includes real-time implementation of the method and testing for prosthesis control.

show abstract

“…As an aggregation criterion, we relied on well-acknowledged feature sets, representing a standard benchmark in a field where TD features proved to be highly effective for classification and found widespread employment, i.e. myoelectric-based gesture recognition [32], [38]. Hence, among 11 of the most commonly used feature sets for sEMG-based gesture classification [32], those made by TD metrics were selected: r Hudgins's feature set (HUDFS) [27]: MAV, WL, ZC, and SSC.…”

Section: Feature Extractionmentioning

confidence: 99%

Identification of Neurodegenerative Diseases From Gait Rhythm Through Time Domain and Time-Dependent Spectral Descriptors

Mengarelli

Tigrini

Fioretti

et al. 2022

IEEE J. Biomed. Health Inform.

View full text Add to dashboard Cite

The analysis of gait rhythm by pattern recognition can support the state-of-the-art clinical methods for the identification of neurodegenerative diseases (NDD). In this study, we investigated the use of time domain (TD) and time-dependent spectral features (PSDTD) for detecting NDD sub-types. Also, we proposed two classification pathways for supporting NDD diagnosis, the first one made by a two-step learning phase, whereas the second one encompasses a single learning model. We considered strideto-stride fluctuation data of healthy controls (CN), patients affected by Parkinson's disease (PD), Huntington's disease (HD), and amyotrophic lateral sclerosis (AS). TD feature set provided good results to distinguish between CN and NDDs, while performances lowered for specific NDD identification. PSDTD features boosted the accuracy of each binary identification task. With k-nearest neighbor classifier, the first diagnosis pathway reached 98.76% accuracy to distinguish between CN and NDD and 94.56% accuracy for NDDs sub-types, whereas the second pathway offered an overall accuracy of 94.84% for a 4-class classification task. Outcomes of this study indicate that the use of TD and PSDTD features, simple to extract and with a low computational load, provides reliable results in terms of NDD identification, being also useful for the development of gait rhythm computer-aided NDD detection systems.

show abstract

A long short-term recurrent spatial-temporal fusion for myoelectric pattern recognition

Cited by 21 publications

References 30 publications

Decoding HD-EMG Signals for Myoelectric Control - How Small Can the Analysis Window Size be?

Decoding HD-EMG Signals for Myoelectric Control - How Small Can the Analysis Window Size be?

Spatio-temporal warping for myoelectric control: an offline, feasibility study

Identification of Neurodegenerative Diseases From Gait Rhythm Through Time Domain and Time-Dependent Spectral Descriptors

Contact Info

Product

Resources

About