CC BY
89
DOI 10.14526/2070-4798-2018-13-4-144-149
Phonomyography and electromyography: correlations and differences from amplitude and frequency parameters
Hanno Felder1*, Laura Steffny2, Daniel Friemert2
Olympiastützpunkt Rheinland-Pfalz/Saarland ORCID: orcid.org/0000-0001-6770-9039
Saarbrücken, Germany [email protected]* 2Koblenz University of Applied Sciences, Remagen, Germany [email protected], [email protected]
Abstract: Electromyography is an important measuring instrument for the investigation of clinical and sport medicine issues, which records myoelectrical signals and thus provides information about the neuromuscular activity of the musculature. An alternative to determine the characteristics of the muscles could be the analysis of the sounds and vibrations which the muscle generates during movement. A measurement method for detecting muscle sounds is the phonomyography. Research Materials. The article examines the correlation between the amplitude, the median frequency plus the mean frequency of the phono- and electromyography and the time during maximum voluntary contraction of the M. rectus femoris. In addition, the integral data for the whole measurement time as well as for four specific moments in time are compared. Research methods. scientific-methodical literature analysis, simultaneous measurement of phono- and electromyographie during maximal voluntary contraction, methods of statistical data handling Results. In contrast to the electromyography, phonomyography does not show any correlation with maximal voluntary contraction. Research results of previous studies have been confirmed as well as new findings on the behavior of integrals are discovered.
Keywords: PMG, EMG, fatigue, acoustic myography, muscle sounds, M. rectus femoris.
For citation: Hanno Felder, Laura Steffny, Daniel Friemert. Phonomyography and electromyography: correlations and differences from amplitude and frequency parameters. The Russian Journal of Physical Education and Sport. 2018; 13 (4): 118-123. DOI 10.14526/2070-4798-2018-13-4-144-149
Introduction
An important measuring instrument for the examination of clinical and sports medicine issues is electromyography, which records myoelectrical signals and thus provides information about the neuromuscular activity of the musculature. An alternative to determine the characteristics of the muscles could be the analysis of the sounds and vibrations which the muscle generates during movement.
Phonomyography (PMG) is a measurement method for detecting muscle sounds. It is used to record and quantify oscillations. PMG records frequencies which have their greatest power under 100 Hz [1]. According to Smith, PMG signal consists of 3 phases [2]:
1. A large lateral movement at the initiation of a contraction generated by a non-simultaneous activation of muscle fibers
2. Smaller subsequent lateral oscillations at the resonant frequency of the muscle
3. Dimensional changes of the active fibers.
In contrast to the electromyography, phonomyography is a mechanical system. Due to this difference, the phonomyography may be an additional method for further determination of muscle quantities in morphology and it could allow new findings of various muscle structures.
The current state of science suggests that with increasing fatigue the amplitude of the electromyography increases and there is a left shift of the frequency spectra.
Therefore, the aim of the investigation is to analyse the behavior of the phonomyography based on changes in amplitude and frequency parameters and to test similarities with the electromyography.
Materials and Methods
Six female and eight male volunteers (age: 20 ± 4 years, weight: 69 ± 12 kg) participated in the investigation. All subjects do more than 5 hours sport per week and do not suffer from injuries of the lower extremities.
Fig. 1. Test Setup. Subjects extend the takeoff leg at a 30 ° angle at maximum load
The investigation was realized at the leg extension (PULSE). The takeoff leg of the volunteers perform a maximal voluntary contraction of the M. rectus femoris. The hip ankle was 90° and the knee was 30° extended. The maximum load for the respective volunteer was determined by a continuous increase in weight. The load was hold as long as possible. The volunteers hold on to handholds beside the hip to stabilize the upper body. Further, the volunteers got feedback on their body and leg posture during the measurement and were also motivated. The measurement was repeated after a 2-minute break. EMG and PMG were recorded simultaneously with the software MyoRESEARCH 3.8 (velamed SCIENCE IN MOTION). The sampling rate was 1000 Hz.
The phonomyography was measured with a piezoelectric transducer (Conrad FT-35T-2.9 al-888) which was secured to the electrode measure point of M. rectus femoris (EMG Fibel [3]) by an elastic band. The signal was digitized by an A/D converter (National Instruments NI USB-6210).
The EMG signal was recorded with the measuring system of the company NORAXON USA, consisting of a DTS EMG sensor and the TeleMyo DTS Desk Resk Receiver. The surface electrodes were placed at a distance of approx. 4 cm above and below the piezoelectric transducer. Prior, a proper skin preparation was carried out for reduction of the skin impedance.
The measured origin signals of the amplitude were processed further by Matlab (Mathworks Matlab R2017A). For this purpose, a root-mean-square-filter (RMS) with the window size 1000 has been implemented and the signals have been standardized relative to the maximum, time and amplitudes. The integrals of the amplitude were calculated by summing the data of all subjects. On the one hand, the entire measuring time is considered and on the other hand, the signal is halved to compare the measurement halves. The unit is arbitrary. For the calculation of the frequencies a sliding window with the size 100 was created so that a Fast Fourier transform (FFT) was carried out continuously for 100 data values. For the average frequency the average of the 100 data values was formed. The median frequency is the value of half area of the envelope.
The correlation between the temporal progress of the amplitude as well as the frequencies were investigated by the correlation of Bravais-Pearson. In contrast, the differences were tested with the T-Test for observation pairs.
Results and Discussion
1. Temporal progress of amplitude
Figure 2 shows the temporal progress of the RMS-EMG and RMS-PMG at maximal voluntary contraction for first and second measurement. Both parameters do not have a significant time-correlation during the first measurement (REMG1 = 0.2031, pEMG1 = 0.6995; RPMG1 =-0.7009, pPMG1 = 0.1208). Also, for the second measurement, the RMS-PMG does not show any significant time-correlation (RPMG2 = -0.3790, pPMG1 = 0.4588), but the RMS-EMG strongly correlates significantly (rEMG2 = 0.9811, PEMG2 = 0.0005).
Rodriguez and colleagues also noted a significant increase in EMG amplitude in their study and no significant change in PMG amplitude during muscular fatigue at 80% MVC [1]. In contrast, they were able to see an increase in both amplitudes for lower MVC (20% and 40%). However, the research group of Orizio found a significant decrease of the PMG amplitude at 80% MVC [4]. It should be noted that this difference is caused through the fact that the study of Orizio was performed on M. biceps brachii and not at the M. rectus femoris. Rodriguez quotes a prolongation of the relaxation time of the muscle fiber and consequently a reduction of the pressure waves as the cause of the absence of the increase in PMG amplitude [1]. A reason for the difference between the EMG amplitudes could be the short duration of time as well as to fluctuations in force.
Fig. 2. Comparison of the RMS-EMG and RMS-PMG at maximal voluntary contraction. The values were calculated for the measurement times 0, 20, 40, 60, 80 and 100% of the measuring time. The values between the points were generated by interpolation
2. Temporal progress of frequencies
Fig. 3. Comparison of the median frequencies of EMG and PMG at maximal voluntary contraction. The values were calculated for the measurement times 0, 20, 40, 60, 80 and 100% of the measuring time. The values between the points were generated by interpolation
Figure 3 shows the progress of the median frequencies at maximal voluntary contraction relating to time. The median frequency of EMG shows a significant decrease in both measurements (REMG1 = -0.9901,
pEMG1 = 0.0001; REMG2 =-0.9887, pEMG2 = 0.0002). By contrast, the frequencies of PMG do not have any significant change (RpM&= -0.5835, ppmg1 = 0.2241; R^ =- 0.3415, ppmg2 = 0.5076).
Comparison of mean frequency of emg and pmg for the first measurement
I—
0 10 20 30 40 50 60 70 80 90 100 time [%\
Comparison of mean frequency of emg ana pmg Tor the second measurement
£ 100 "O
B 50
------}......
10 20 30 40 50 60 70 80 90 100 time [K|
Fig. 4. Comparison of the mean frequencies of EMG and PMG at maximal voluntary contraction. The values were calculated for the measurement times 0, 20, 40, 60, 80 and 100% of the measuring time.
The values between the points were generated by interpolation
Figure 4 shows the temporal progress of the mean frequencies. Measurement 1 shows no significant time-correlation for both parameters (REMG1 = -0.3366, pEMG1 = 0.5141; RPMG1 = -0.7917, pPMG1 = 0.0606). The mean frequencies of the PMG for the second measurement also do not correlate significantly (RPMG2 = -0.5800, pPMG2 = 0.2275), whereas a strongly significant correlation could be identified for EMG (rEMG2 = 0.90^ Pemg2 = °.°122).
Dalton and Stokes were able to notice a square increase in the mean frequency with increasing force in the study of isometric contractions for various MVC values, but also there was no significant difference between a rested and a fatigued muscle [5]. The reason for the difference between the mean frequencies of EMG between the first and second measurement is barely discussed in the literature. Probably, there is a migration of muscle activity between synergist and a reduction of co-activation of antagonists [3].
3. Integral
The integrals of EMG and PMG for the entire duration of measurement were investigates as a further parameter (figure 5). The average value of EMG is 86.4 ± 4.9 a.u. and that of the PMG 80.58 ± 8.0 a.u.. It is assumed that the mean results of the integrals are significantly different due to the result of the t-test (H = 1, p = 1, 7e-10).
Fig. 5. Comparison of the mean values of the integrals of EMG and PMG over the entire duration
of measurement
In addition, figure 6 compares the integrals of the first and second half of the signal. For the first half, the average value of the EMG integral is 41,87 ± 3.63 a.u. and the PMG integral is 40.98 ± 3.78 a.u.. There are no significant differences based on the t-test (h = 0, p = 0.4254). The EMG integral of the second measuring half is 44.54 ± 2.83 a.u. and the PMG Integral is 39.60 ± 5.50 a.u.. According to the t-test, the mean values are significantly different at this time (h = 1, p < 0.01). The t-test suggests for EMG a significant
difference between the two measuring halves (h = 1, p = 0.0028). In contrast, the average values of the PMG do not differ significantly (h = 0, p = 0.1516).
Fig. 6. Comparison of the mean values of the integrals of EMG and PMG at different measuring
dates
Several research groups, including Dalton and Stokes, discovered a linear increase in integrals with increasing force. The integral of PMG is up to an MVC of 60% higher in percentage than that of EMG. From 80% MVC, they also do not differ significantly from each other [6]. These results are different from the results shown in figure 5. This could be due to the change in the integrals with increasing fatigue, because they behave in the opposite way. In addition, the increase of the EMG integral is higher than the decrease of the PMG integral, so that the total integral of EMG is higher.
Conclusion
The study did not find a significant correlation for the PMG between the parameters and the time at maximal voluntary contraction. Only the EMG showed an increase in amplitude as well as a decrease of the median frequency. A possible indicator for quantifying muscular fatigue is the comparison of the integrals, because the integral of the PMG is higher than that of the EMG at low MVC values and the integrals are the same at high values for rested muscles. As fatigue increases, the relations may shift due to the fact that the integrals of the PMG decrease and the integrals of the EMG increase.
PMG records a complex signal whose investigation is still in an initial phase. For these reasons, it has not yet been sufficiently investigated which factors influences the signal and how big their amount is. Dalton and Stokes mention a combination of factors. These include, among other things, the physical properties and stiffness of the muscles, the motor-unit-frequency, the density and elasticity of tissues as well as the body temperature [5].
In order to assess the potential of the phonomyography, further studies are necessary to quantify the influence of various factors and to analyse the behavior of the phonomyography under various conditions and parameters.
References
1. Rodriguez A. A., Agre, J.C., Knudtson E. R., Franke T.M., Ng A. V. Acoustic myography compared to electromyography during isometric fatigue and recory. Muscle & nerve. 1993; 16(2): 188-192. DOI: 10.1002/mus.880160212
2. Smith D.B., Housh T.J., Johnson G.O., Evetovich T.K., Ebersole K.T., Perry S.R. Mechanomyographie and electromyographie responses to eccentric and concentric isokenetic muscle actions of the biceps brachii. Muscle & nerve. 1998; 21 (11): 1438-1444.
3. Konrad P. EMG-Fibel. Eine praxisorientierte Einführung in die kinesiologische Elektromyographie.
2011.
4. Orizio C., Perinini R., Veicsteinas A. Changes of muscular sound during sustained isometric contraction up to exhaustion. In: Journal of Applied Physics. 1989; 66:1593-1598.
5. Dalton P. A., Stokes M. J. Frequency of acoustic myography during isometric contraction of fresh and fatigued muscle and during dynamic contractions. Muscle & nerve. 1993; 16 (3): 255-261. DOI: 10.1002/mus.880160303
6. Dalton P. A., Stokes M. J. Acoustic myographix activity increases linearly up to maximal voluntary isometric force in the human quadriceps muscle. In: Journal of the Neurological Sciences. 1990: 163-167.
Submitted: 01.10.2018
Hanno Felder* - Doctor of Sports Science, Professor, Olympiastutzpunkt Rheinland-Pfalz/Saarland, Hermann-Neuberger-Sportschule 2, Saabrucken, Germany, 66123, e-mail: h.felder@olympiastuetzpunkt. org
Laura Steffny - Bachelor of Science, Koblenz University of Applied Sciences, Joseph-Rovan-Allee 2, Remagen, Germany, 53424, e-mail: la.steffny@ live.de
Daniel Friemert - Master of Science, Substitute Professor, Koblenz University of Applied Sciences, Joseph-Rovan-Allee 2, Remagen, Germany, 53424, e-mail: [email protected]
