A method to test reliability and accuracy of the decomposition of multi-channel long-term intramuscular EMG signal recordings

Zennaro, Daniel; Läubli, Thomas; Wellig, Peter; Krebs, Denise; Schnoz, Michael; Klipstein, Andreas; Krueger, Helmut

doi:10.1016/s0169-8141(02)00126-9

Cited by 15 publications

(12 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fifteen percent of the MUAPs of the 15 MU pairs were affected by observed variations in amplitude of the same MU (±28% of the maximum peak-to-peak amplitude). Similar results were found in a previous study (Zennaro et al 2002), where amplitude modulations in long-term recordings (10 min long) of more than ±30% of the maximum peak-to-peak amplitude were observed. Although slow shape changes over time are correctly detected by the tracking algorithm, abrupt changes are often misclassified as newly recruited MUs.…”

Section: Accuracy Of the Decomposition Results (Concurrently Active Mus)supporting

confidence: 92%

“…To test the accuracy of the decomposition program, artificially generated signals with known features were used as a reference (Zennaro et al 2002). For details on the procedure for obtaining artificially generated intramuscular EMG signals, refer to Farina et al (2001).…”

Section: Methods To Test Accuracy and Reliabilitymentioning

confidence: 99%

“…The stored intramuscular EMG signals were decomposed into individual MUAP trains using the automatic decomposition software EMG-LODEC, as reported in detail by Zennaro et al (2002Zennaro et al ( , 2003. The program uses wavelet coefficients, template matching, template updating and superposition resolution to identify the individual firing times of the MUs.…”

Section: Data Analysis and Statisticsmentioning

confidence: 99%

“…The problem is that to understand the physiological mechanisms behind the development of muscle fatigue, pain, and strain at lowintensity work, EMG recordings from a minute up to several hours duration must be accurately decomposed. For this reason we developed EMG-LODEC (ElectroMyoGram LOng-term DEComposition), an automatic decomposition software program (Zennaro et al 2002(Zennaro et al , 2003. In contrast to existing decomposition methods, the decomposition software EMG-LODEC is especially designed for multi-channel long-term recordings of signals of slight muscle movements.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Motor unit identification in two neighboring recording positions of the human trapezius muscle during prolonged computer work

Zennaro¹,

Läubli²,

Krueger³

2003

European Journal of Applied Physiology

Self Cite

View full text Add to dashboard Cite

Work-related shoulder-neck pain is a major health risk in computer operators. To understand the physiological mechanisms behind the development of these disorders, EMG recordings of some minutes up to several hours must be accurately decomposed. For this reason we developed EMG-LODEC, an automatic decomposition software program, especially designed for multi-channel long-term recordings of signals detected during slight muscle movements. The subjects executed a 30-min computer task to simulate real work conditions while working at an ergonomically designed workstation. Six-channel intramuscular EMG signals were recorded from two positions of the upper trapezius muscle. The EMG signals were decomposed into individual motor unit action potential trains using EMG-LODEC. The study design enabled us first to study the dependence of intramuscular analysis on the insertion points and second to test the accuracy of the decomposition technique under laboratory conditions during a real experiment. The two positions yielded 887 motor units -452 located in position 1 and 435 in position 2. Although the numbers of detected action potentials were strongly correlated between the two insertion positions, different motor units were mostly recorded. In particular, the detection of continuously active motor units is specific for the selected insertion points and may not be representative of a muscle, not even for parts with common functions. The approach for the quantitative evaluation of the decomposition technique was to independently decompose two signals that were simultaneously detected by separate sets of wire electrodes placed close to each other in the muscle. Common trains discovered in each signal were compared for consistency. A cross-correlation analysis was performed to find corresponding motor unit pairs that were concurrently active. Concurrently active motor units were found in six subjects. For these motor units the extent of simultaneous occurrence of motor unit action potentials between the two positions ranged from 23% to 78% depending on the distinction of the single motor units and the number of superimposed motor unit action potentials. High concordance was seen in 3 out of the 15 motor unit pairs. Based on the results, EMG-LODEC is capable of providing reliable decompositions with satisfying accuracy and reasonable processing time. EMG-LODEC is suitable for the study of motor unit discharge patterns and recruitment order in subjects with and without musculoskeletal pain during long-term measurements to study work-related musculoskeletal disorders.

show abstract

Section: Accuracy Of the Decomposition Results (Concurrently Active Mus)supporting

confidence: 92%

Section: Methods To Test Accuracy and Reliabilitymentioning

confidence: 99%

Section: Data Analysis and Statisticsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Motor unit identification in two neighboring recording positions of the human trapezius muscle during prolonged computer work

Zennaro¹,

Läubli²,

Krueger³

2003

European Journal of Applied Physiology

Self Cite

View full text Add to dashboard Cite

show abstract

“…Most of the clustering techniques developed for EMG decomposition are based on adaptations of general clustering algorithms such as the nearest-neighbor [1], [26], [53], [115], single linkage [54], [55], [91], [93], [94], [116], [117], K-means [1], [7], [26], [68], [74], fuzzy c-means [66], [67], [118], minimal spanning tree [54][55][56], [59], [65], [69], [119], [120], leader-based clustering [6], [25], and self-organizing neural nets algorithm [58]. In many of these algorithms MU firing pattern information is used passively or actively along with MUP shape information to assign an individual MUP to the correct train.…”

Section: Clustering and Supervised Classification Of Detected Mupsmentioning

confidence: 99%

EMG Signal Decomposition Using Motor Unit Potential Train Validity

Parsaei

Stashuk

2013

IEEE Trans. Neural Syst. Rehabil. Eng.

View full text Add to dashboard Cite

Electromyographic (EMG) Decomposition

Marateb

McGill

2016

Wiley Encyclopedia of Electrical and Electronics Engineering

View full text Add to dashboard Cite

An electromyographic (EMG) signal is a biomedical signal derived from measuring electrical potentials generated in contracting muscles. EMG signals can be recorded using either intramuscular (iEMG) or surface electrodes (sEMG). EMG signals are composed of motor unit action potential (MUAP) trains, the MUAP being the signal detected in response to the discharge of a group of muscle fibers innervated by a single motoneuron. The process of extracting the MUAP trains from an EMG signal is referred to as EMG decomposition. EMG decomposition makes it possible to study the behavior of individual motoneurons and thus to decode the neural drive to a muscle at a very precise level. This article reviews the different electrodes used to record iEMG and sEMG signals, the main steps in pattern‐matching decomposition algorithms (filtering, segmentation, clustering, classification, and resolving superpositions), the blind‐source separation techniques used for decomposing high‐density sEMG signals, the principal decomposition programs that have been described in the literature, including those that are publicly available, and approaches for assessing the accuracy of decomposition algorithms and particular decomposition results. Finally, an interactive tutorial on iEMG decomposition is provided.

show abstract

A method to test reliability and accuracy of the decomposition of multi-channel long-term intramuscular EMG signal recordings

Cited by 15 publications

References 18 publications

Motor unit identification in two neighboring recording positions of the human trapezius muscle during prolonged computer work

Motor unit identification in two neighboring recording positions of the human trapezius muscle during prolonged computer work

EMG Signal Decomposition Using Motor Unit Potential Train Validity

Electromyographic (EMG) Decomposition

Contact Info

Product

Resources

About