The EMG Signal

1
The EMG Signal

• EMG Frequency Spectrum
• Fatigue
• Signal Processing.4

2
Motor Unit Firing Rates

• Firing rate frequency
• No. of cycles (firings) per unit of time
• Example 175 cps 175 Hertz (Hz)
• Range of frequencies the (Power) Spectrum
  the Bandwidth
• Slow twitch motor units (tonic - Type I)
• Frequency range (20) 70 - 125 Hz
• Fast twitch motor units (phasic - Type II)
• Frequency range 126 - 250 Hz

3
The Power Spectrum
ST Slow twitch mus FT Fast twitch mus
ST FT
Bandwidth
4
Muscle Fatigue.1

• Grossly manifests as a decrease in tension/force
  (and power) production
• Insufficient O2
• Energy stores used up/exhausted
• Lactic acid builds up
• Circulatory system has difficulty removing lactic
  acid
• Accumulates in extracellular fluid surrounding
  muscle fibers (Bass Moore, 1973 Tasaki et al.,
  1967)
• Decreases pH

5
Muscle Fatigue.2

• Decreased pH causes a decrease in the conduction
  velocity of muscle fibers
• Fast twitch (phasic) motor units relying on
  anaerobic respiration will be more sensitive to
  circulatory inefficiency and will decrease their
  activity or stop functioning before slow twitch
  (tonic - aerobic) motor units (De Luca et al.,
  1986)

6
Muscle Fatigue.3

• Sustained muscle contractions (i.e., isometric)
  may cause local occlusion of arterioles due to
  internal pressure and have a similar limiting
  effect on circulation with resultant decrease in
  extracellular pH (De Luca et al., 1986)

7
Muscle Fatigue.4

• With decreased conduction velocity of muscle
  fibers
• Decrease in peak twitch tensions
• Increase in contraction times
• Corresponding decrease in firing frequency
• The result is a decrease in force

8
Muscle Fatigue.5

• With fatigue there is a change in the shape of
  action potentials (Enoka, 1994)
• Decreased amplitude
• Increased duration
• Result is a EMG spectrum shift to lower
  frequencies (Winter, 1990)

9
Muscle Fatigue.6

• As fatigue progresses there is a shift to lower
  frequencies
• Fast twitch (higher frequency) motor units drop
  out first
• Slow twitch (lower frequency) motor units
  retained

10
Muscle Fatigue.7

• Therefore a spectral shift to the left

11
Spectral Analysis

• Indicies of frequency shift (Soderberg Knutson,
  2000)
• Mean power frequency
• Median power frequency
• More commonly used
• Not susceptible to extremes in the range
  (bandwidth)
• Therefore a more sensitive measure (Knaflitz De
  Luca, 1990)
• Therefore a decrease in the median power
  frequency serves as an index of fatigue

12
Frequency-Domain Analysis.1

• Transformation from the time domain to the
  frequency domain
• Fast Fourier Transformation (FFT)
• Fourier series of equations

13
Frequency-Domain Analysis.2

• Removes the time between successive action
  potentials so that they appear as periodic
  functions of time

Pre-fatigue
Fatigue
14
Frequency-Domain Analysis.3

• Action potentials represented by a best-fitting
  combination of sine-cosine functions to
  characterize the frequency and amplitude of the
  signal
• Result is a single line (per frequency)

Pre-fatigue
Fatigue
15
Frequency-Domain Analysis.4

• Result is plotted on a frequency-amplitude graph

16
Frequency-Domain Analysis.5

• Major factors that cause an active change in
  frequency
• Action potential shape (see above)
• Decrease motor unit discharge rate

17
Frequency-Domain Analysis.6

• Action potential shape
• Changes due to conduction velocity rate along
  sarcolema of muscle fiber
• As conduction velocity decreases the duration of
  action potential decreases causing a decrease in
  the median power frequency (De Luca, 1984)

• Decrease in motor unit discharge rate
• Causes grouping of action potentiasl at low
  frequencies 10 Hz (Krogh-Lund Jogensen, 1992)

18
Frequency-Domain Analysis.7

• Outcome a decrease in median power frequency

Shift to the left
19
Frequency-Domain Analysis.8

• Converse relationship with increasing force
  production
• Moritani Muro (1987) found a significant
  increase in mean power frequency with increasing
  force during an MVC of the biceps brachii

20
Median Power Frequency Calculation Procedure

• Sample data in multiples of x2 (Example 1024 Hz)

21
Median Power Frequency Calculation Procedure

• Sample data in multiples of x2 (Example 1024 Hz)
• Rectify and filter (BP or LP) raw signal

22
Median Power Frequency Calculation Procedure

• Sample data at multiples of x2 (Example 1024 Hz)
• Rectify and filter (BP or LP) raw signal
• Apply FFT

Hz
23
Median Power Frequency Calculation Procedure

• Sample data at multiples of x2 (Example 1024 Hz)
• Rectify and filter (BP or LP) raw signal
• Apply FFT
• Compute median (or mean power) frequency

24
Spec_rev with cursors.vi (with BP filter cutoffs
20 500 Hz)
25
Reference Sources

• Bass, L., Moore, W.J. (1973). The role of
  protons in nerve conduction. Progressive
  Biophysics and Molecular Biology, 27, 143.
• Bracewell, R.N. (1989). The Fourier transform.
  Scientific American, June, 86-95.

26
Reference Sources

• De Luca, C. J. (1984). Myoelectric manifestations
  of localized muscular fatigue in humans. CRC
  critical reviews in biomedical engineering, 11,
  251-279.
• De Luca, C.J., Sabbahi, M.A., Stulen, F.B.,
  Bilotto, G. (1982). Some properties of median
  nerve frequency of the myoelectric signal during
  localized muscular fatigue. Proceedings of the
  5th International Symposium on Biochemistry and
  Exercise, 175-186.
• Enoka, R. M. (1994). Neuromechanical basis of
  kinesiology (Ed. 2). Champaign, Ill Human
  Kinetics, pp. 166-170.

27
Reference Sources

• Fahy, K., Pérez, E. (1993). Fast Fourier
  transforms and the power spectra in LabVIEW.
  Application Note 040, February, Austin TX
  National Instruments Corp. (www.ni.com) (pn
  340479-01)
• Gniewek, M.T. (19xx). Waveform analysis using the
  Fourier transform. Application Note-11, Great
  Britain AT/MCA CODAS-Keithly Instruments, Ltd.,
  pp1-6.

28
Reference Sources

• Harvey, A.F., Cerna, M. (1993). The
  fundamentals of FFT-based signal analysis and
  measurements in LabVIEW and LabWindows.
  Application Note 041, November, Austin, TX
  National Instruments Corp. (www.ni.com) (pn
  340555-01.
• Krogh-Lund, C., Jorgensen, K. (1992).
  Modification of myo-electric power spectrum in
  fatgiue from 15 maximal voluntary contraction of
  human elbow flexor muscles, to limit of
  endurance reflection of conduction velocity
  variation and/or centrally mediated mechanisms?
  European Journal of Applied Physiology, 64,
  359-370.

29
Reference Sources

• Moritani, T., Muro, M. (1987). Motor unit
  activity and surface electromyogram power
  spectrum during increasing force of contraction.
  European Journal of Applied Physiology, 56,
  260-265.
• Merleti, R., Knaflitz, M., De Luca, C.J.
  (1990). Myoelectric manifestations of fatigue in
  voluntary and electrically elicited contractions.
  Journal of Applied Physiology, 69, 1810-1820.

30
Reference Sources

• Ramirez, R.W. (1985). The FFT fundamentals and
  concepts. Englewood Cliffs, NJ Prentice Hall
  PTR.
• Soderberg, G.L., Knutson, L.M. (2000). A guide
  for use and interpretation of kinesiologic
  electromyographic data. Physical Therapy, 80,
  485-498.
• Tasaki, I., Singer, I., Takenaka, T. (1967).
  Effects of internal and external ionic
  environment on the excitability of squid giant
  axon. Journal of General Physiology, 48, 1095.

31
Reference Sources

• Weir, J.P., McDonough, A.L., Hill, V. (1996).
  The effects of joint angle on electromyographic
  indices of fatigue. European Journal of Applied
  Physiology and Occupational Physiology, 73,
  387-392
• Winter, D.A. (1990). Biomechanics and motor
  control of human movement (2nd Ed). New York
  John Wiley Sons, Inc., 191-212.

32
(No Transcript)