Common misunderstandings about studying motor units

Jim Richards

Myth There is no need for motor unit data: surface EMG gives me all the information I need.


“Surface EMG is like listening to a symphony; you get a general sense of the music from all of the instruments combined. Decomposing sEMG into motor unit firings is like listening to the same symphony, while distinguishing all of the individual instruments from the conductor’s score.”

Jim Richards, PhD
Professor – Biomechanics, University of Central Lancashire

Jason DeFreitas

Myth The study of motor unit behavior has no real-world applications.


“The analysis of motor unit behavior provides a fundamental perspective to understanding how muscles respond to exercise, clinical interventions, injury and disease. In fact, in just the past few years, research has shown:

  • motor unit-based parameters may contribute to explain muscle weakness in stroke
  • adaptations in motor unit behavior explain the mechanism of muscle fatigue development
  • motor unit-based parameters may provide a non-invasive biomarker of muscle hypertrophy and atrophy
  • motor unit parameters can help explain motor deficits with aging, including contributions to risk of falling
  • motor unit behavior have shown a response to proprioceptive interventions for knee pain

Jason DeFreitas, PhD
Assistant Professor – Exercise Physiology, Oklahoma State University

Xiaogang Hu

Myth Studying motor units is too complex & time-consuming.


“With today’s technology, motor unit studies are within everyone’s reach! For instance, dEMG gives you easy-to-apply sensors and easy-to-use automatic analysis tools that let you measure all aspects of motor unit behavior, even if you are not a signal processing expert.”

Xiaogang Hu, PhD
Assistant Professor – Biomedical Engineering; Director – Laboratory of Neuromechanics, UNC Chapel Hill

Myth More sEMG channels = more motor units.


“I think the biggest floating myth is that more sEMG channels always give more motor units. The empirical data has consistently shown that just 4 high-fidelity sEMG signals provide a greater number of motor units than larger sensors. It is the quality of the sEMG signal not the quantity that matters.”

Zev Rymer, MD PhD
Director – Sensor Motor Performance Program, Shirley Ryan AbilityLab

Myth All EMG decomposition algorithms are the same.


“Anyone can stitch together generic machine learning code and say they have a new decomposition algorithm. But how accurate is it? It takes decades of research and engineering effort to solve this challenge in a way that provides reliable measurements of the action potentials and firing behavior of motor units. See beyond the marketing hype; ensure that your research tools are extensively validated by independent reviewers and backed by years of peer-reviewed investigations.”

Joshua Kline, PhD
Lead Research Engineer, Delsys Inc.