Multi-channel intramuscular and surface EMG decomposition by convolutive blind source separation
Francesco Negro1, Silvia Muceli, Anna Margherita Castronovo
1Institute of Neurorehabilitation Systems, Bernstein Focus Neurotechnology Göttingen, Bernstein Center for Computational Neuroscience, University Medical Center Göttingen, Georg-August University of Göttingen, Göttingen, Germany.
Journal of Neural Engineering
|March 1, 2016
Summary
This study introduces a new framework for analyzing muscle electrical activity (EMG) signals. The method accurately decomposes multi-channel signals, enabling the study of numerous motor units simultaneously.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Physiology
Background:
- Motor unit behavior is traditionally studied using selective recording systems and decomposition algorithms for electromyography (EMG) signals.
- Existing methods face challenges in discriminating individual motor unit action potentials from complex multi-unit signals.
Purpose of the Study:
- To present a general framework for decomposing multi-channel intramuscular and surface EMG signals.
- To extensively validate this novel decomposition approach using experimental recordings.
Main Methods:
- Described conditions for convolutive blind separation model assumptions.
- Proposed an iterative source extraction approach following convolutive sphering.
- Validated the method on intramuscular and high-density surface EMG signals from human muscles.
Main Results:
- Identified an average of 14 common sources with 92.8% agreement in discharge timings.
- Achieved a Decomposability Index of 16.0 ± 2.2 for automatically decomposed signals, comparable to manual decomposition (15.0 ± 3.0).
Conclusions:
- The developed method offers a robust framework for decomposing both invasive and non-invasive EMG signals.
- This approach facilitates the investigation of a large number of concurrently active motor units, advancing motor control research.
