Development of sEMG sensors and algorithms for silent speech recognition

Meltzner, Geoffrey S.; Heaton, James T.; Deng, Yunbin; Luca, Gianluca De; Roy, Serge H.; Kline, Joshua C.

doi:10.1088/1741-2552/aac965

Cited by 76 publications

(55 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Three time domain and three frequency domain features were computed from every EMG contraction segment in order to analyze the differences in amplitude and frequency of the signals acquired by CREs and CDEs. Several previous studies used these features in sEMG recordings of the muscles used to swallow [ 11 , 40 , 41 ]. Root mean square (RMS) is equivalent to the square root of the variance of the signal with zero mean and is a typical measure used to assess signal power in time domain.…”

Section: Methodsmentioning

confidence: 99%

Evaluation of Swallowing Related Muscle Activity by Means of Concentric Ring Electrodes

Garcia-Casado

Prats-Boluda

Ye-Lin

et al. 2020

Sensors

View full text Add to dashboard Cite

Surface electromyography (sEMG) can be helpful for evaluating swallowing related muscle activity. Conventional recordings with disc electrodes suffer from significant crosstalk from adjacent muscles and electrode-to-muscle fiber orientation problems, while concentric ring electrodes (CREs) offer enhanced spatial selectivity and axial isotropy. The aim of this work was to evaluate CRE performance in sEMG recordings of the swallowing muscles. Bipolar recordings were taken from 21 healthy young volunteers when swallowing saliva, water and yogurt, first with a conventional disc and then with a CRE. The signals were characterized by the root-mean-square amplitude, signal-to-noise ratio, myopulse, zero-crossings, median frequency, bandwidth and bilateral muscle cross-correlations. The results showed that CREs have advantages in the sEMG analysis of swallowing muscles, including enhanced spatial selectivity and the associated reduction in crosstalk, the ability to pick up a wider range of EMG frequency components and easier electrode placement thanks to its radial symmetry. However, technical changes are recommended in the future to ensure that the lower CRE signal amplitude does not significantly affect its quality. CREs show great potential for improving the clinical monitoring and evaluation of swallowing muscle activity. Future work on pathological subjects will assess the possible advantages of CREs in dysphagia monitoring and diagnosis.

show abstract

Section: Methodsmentioning

confidence: 99%

Evaluation of Swallowing Related Muscle Activity by Means of Concentric Ring Electrodes

Garcia-Casado

Prats-Boluda

Ye-Lin

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…Using (14), there are two types of measurement indicators that can be used to identify the best muscle location during the recitation of Ruqyah. The first indicator is the amount of information embedded in an EMG signal.…”

Section: Resultsmentioning

confidence: 99%

“…Finally, all signals are normalized so that the amplitude between the peaks is equal to 1 [35]. All clean signals are then processed for the measurement of the , calculated using (14) as described in the previous section.…”

Section: Methodsmentioning

confidence: 99%

“…Several areas have been widely studied, which include: a) Specific applications; like in prosthesis control [10], oral motor treatment for speech production [11] and improve communication while subject wears a self-contained breathing apparatus (SCBA) [12]. b) Feature extraction and classification techniques [13][14][15][16] c) Language or Dialect variation [17][18][19] d) Speech synthesis [20][21][22] This study will look into another equally important factor in EMG-based speech recognition technology, namely the impact of muscles' selection towards the quality of speech. The speech production begins in the lungs as illustrated in Ture 1.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Quality assessment on muscle locations for speech representation

Safie

Rahim

2020

IJEECS

View full text Add to dashboard Cite

<span lang="EN-US">There are more than 68 muscles, which are activated either simultaneously or sequentially during speech production. To monitor the signals from all these muscles at once, involve a lot of sensors and such system is very expensive. In the Quran therapeutic treatment applications, the use of specific muscles is very important, for the production of correct Arabic pronunciation. The proper pronunciation will improve the reader's understanding of what is being read, thus assisting the effectiveness of the therapy process. The objective of this study is to identify the most optimal muscle location, which is suitable for monitoring the quality of a recitation during the Quran’s therapeutic process, based on the information content embedded in their Electromyogram (EMG) signals. Empirical Mode Decomposition (EMD) technique was used in this study to extract features of the EMG while the combination of Hilbert Huang Spectral Entropy (HHSE) and Kullback Leibler Divergence (KLD) techniques were used to quantify the information content in each feature. Combination of these techniques managed to rank ten widely used speech muscles in the literature based upon their information content. Four muscle locations have been suggested, which is believed to be sufficient in developing a low-cost self-assessment system for monitoring Quran recitation.</span>

show abstract

“…The second sEMG-based silent speech recognition device has been described by a team at Altec Inc. (Boston, Massachusetts) and was partially funded by DARPA and an NIDCD grants. 31 It has been developed for silent articulation for laryngectomees and for subvocal speech allowing direct communication with computer devices. 32 It offers the most advanced algorithmic and hardware system, but with crucial design differences with our approach.…”

Section: Discussionmentioning

confidence: 99%

Pilot study for a novel and personalized voice restoration device for patients with laryngectomy

Rameau¹

2019

Head & Neck

View full text Add to dashboard Cite

Background: The main modalities for voice restoration after laryngectomy are the electrolarynx, and the tracheoesophageal puncture [Correction added on 30 January 2020 after first online publication: The preceding sentence has been revised. It originally read "The main modalities for voice restoration after laryngectomy are the electrolarynx and the tracheoesophageal puncture."]. All have limitations and new technologies may offer innovative alternatives via silent speech. Objective: To describe a novel and personalized method of voice restoration using machine learning applied to electromyographic signal from articulatory muscles for the recognition of silent speech in a patient with total laryngectomy.Methods: Surface electromyographic (sEMG) signals of articulatory muscles were recorded from the face and neck of a patient with total laryngectomy who was articulating words silently. These sEMG signals were then used for automatic speech recognition via machine learning. Sensor placement was tailored to the patient's unique anatomy, following radiation and surgery. A personalized wearable mask covering the sensors was designed using 3D scanning and 3D printing.Results: Using seven sEMG sensors on the patient's face and neck and two grounding electrodes, we recorded EMG data while he was mouthing "Tedd" and "Ed." With data from 75 utterances for each of these words, we discriminated the sEMG signal with 86.4% accuracy using an XGBoost machinelearning model. Conclusions:This pilot study demonstrates the feasibility of sEMG-based alaryngeal speech recognition, using tailored sensor placement and a personalized wearable device. Further refinement of this approach could allow translation of silently articulated speech into a synthesized voiced speech via portable devices.

show abstract

Development of sEMG sensors and algorithms for silent speech recognition

Cited by 76 publications

References 23 publications

Evaluation of Swallowing Related Muscle Activity by Means of Concentric Ring Electrodes

Evaluation of Swallowing Related Muscle Activity by Means of Concentric Ring Electrodes

Quality assessment on muscle locations for speech representation

Pilot study for a novel and personalized voice restoration device for patients with laryngectomy

Contact Info

Product

Resources

About