Surface Electromyography and Muscle Fatigue

Changes in Frequency Content with Muscle Fatigue

Investigator:  Eric Alan Beltt

Advisor:  Professor William K. Durfee

School:  University of Minnesota

Course:  Methods of Experimental Physics II

Objective:

To confirm that the characteristic frequency content of an EMG signal changes as a muscle fatigues. To gain engineering experience through the design and construction of the EMG apparatus.

Background:

In order to flex a muscle, an electrical signal has to travel across the fibers that make up that muscle. That small electric signal can be detected through electrodes placed on the surface of the skin. With some filtering and amplification, along with careful circuit and electrode design, that signal can be transformed into useful information about muscle activity. The picture below (Delsys) is an EMG signal in the time domain, with the measured muscle slowly contracting.

As a muscle fatigues, the individual fibers that make up that muscle slowly stop working. As a result, the mean frequency of the EMG signal slowly shifts to lower frequencies. This is an important phenomenon because it can be used as a fatigue index. The objective of this experiment was to observe that shift.

The subject of the experiment held an appropriately sized barbell so that the bicep was fatigued within 2 minutes. The subject held the barbell so that the elbow was at a right angle, and the forearm was parallel to the ground. I recorded half-second epochs of data for two minutes, at a sampling rate of 4096Hz. Finally, I calculated the mean frequency for each epoch of data using the formula,

,

where A(f) was the amplitude at the frequency f.

Results:

The blue dots above are a 100Hz baseline sinusoidal signal with an amplitude of 10uV. The red dots are the data I recorded from my right bicep. A linear least squares fit yielded a slope of -0.15 +/- .01 Hz/s, confirming that I actually did observe the predicted shift in mean EMG frequency.

Miscellaneous EMG Signals:

These are signals I recorded from my jaw and right bicep (2x) plotted in the time and frequency domains respectively. Note that the time domain signal is on a much shorter scale than the sample shown in the background section.

Links:

Introduction to EMG: Tutorials and WhitePapers

Various EMG Signals