Related Experiment Videos
Decomposition of surface EMG signals
Carlo J De Luca1, Alexander Adam, Robert Wotiz
1NeuroMuscular Research Center, Boston, MA 02215, USA. cjd@bu.edu
Journal of Neurophysiology
|August 11, 2006
Summary
This study introduces an AI-driven technique to decompose surface electromyographic (sEMG) signals into motor unit (MU) action potential trains. The method achieves high accuracy, enabling new research into MU behavior in difficult-to-study muscles.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Signal Processing
Background:
- Surface electromyography (sEMG) is a non-invasive method to study muscle activity.
- Decomposing sEMG signals into individual motor unit (MU) action potential trains is challenging but crucial for detailed muscle analysis.
- Existing methods often lack the precision to accurately identify and separate individual MU signals.
Purpose of the Study:
- To present an early version of an Artificial Intelligence (AI) based technique for decomposing sEMG signals into MU action potential trains.
- To assess the accuracy and yield of this novel decomposition method.
- To explore the potential of this technique for studying MU behavior in muscles not easily accessible with traditional needle sensors.
Main Methods:
- Utilized a surface sensor array to collect four channels of differentially amplified sEMG signals.
- Developed a knowledge-based AI framework with algorithms for automatic sEMG decomposition.
- Employed an Interactive Editor to refine decomposition accuracy.
- Validated accuracy by comparing surface-detected MU firings with simultaneous needle sensor recordings.
Main Results:
- Automatic decomposition accuracy ranged from 75% to 91%.
- Interactive editing improved accuracy to over 97% for 30-second epochs.
- Successfully decomposed up to six MU action potential trains from orbicularis oculi, platysma, and tibialis anterior muscles.
- Observed a low yield, typically 5 or fewer MUs per contraction, with potential for improvement.
Conclusions:
- The AI-based sEMG decomposition technique shows promise for analyzing MU behavior, particularly in muscles difficult to study with needle electrodes.
- The inverse relationship between recruitment threshold and firing rate, known for spinal nerve-innervated muscles, was also observed in cranial nerve-innervated orbicularis oculi and platysma.
- Orbicularis oculi and platysma exhibited higher and more widespread firing rates compared to large limb muscles, suggesting unique physiological characteristics.