Analysis of skeletal muscle performance using piezoelectric film sensors

Abstract

A flexible piezoelectric thin film sensor has been proposed recently in several studies for detection of muscle movements. The objective of this study was to investigate the ability of this sensor to assess skeletal muscle performance and fatigue under isokinetic contractions. Simultaneous noninvasive measurements of muscles activity were done using surface electromyography (EMG) electrodes and two thin film piezoelectric sensors. Measurements were taken from the biceps during slow and fast elbow flexion with and without strong grip, during different weight lifting and from the gastrocnemius during treadmill marching at speeds of 4 and of 10 kph. The results shows correlation between the onset of EMG and the piezoelectric sensors (Piezo) signals during muscle contraction. Increasing contraction intensity increase significantly both EMG and Piezo signals. Higher contractions velocity increased Piezo signal. Opposite linear relation was found between the average maximal EMG envelope amplitudes and the average maximal Piezo peaks with increasing loads. The significant decrease in the maximal Piezo peaks with time of all 3 subjects during elbow flexion while holding weight suggests the ability of piezoelectric thin film sensor to track muscle fatigue during isokinetic contractions.

Keywords

Electromyography isokinetic biceps contractions fatigue

1. Introduction

Several methods are being used to assess skeletal muscle activity as electromyography (EMG), mechanomyography (MMG) and ultrasound imaging [1, 2, 3]. Surface EMG is common and reliable technique that detects superimposed motor unit action potentials from fibers in the area of the detecting electrodes [1]. EMG has been used extensively for assessing skeletal muscle activity, to detect human movements and muscle fatigue [4, 5, 6]. Sonomygraphy and MMG are other techniques for the evaluation of skeletal muscles function [7, 8, 9]. Sonomygraphy uses ultrasound to measure muscle morphological changes during static and dynamic contraction [10]. MMG signal is being measured at the skin surface using an accelerometer sensor, piezoelectric contact sensor or condenser microphone [8, 11] to detect the dimensional changes of the active muscle fibers. MMG measurements were done to detect neuromuscular disorders [12, 13, 14, 15], evaluate muscle mechanical activity [16], assess muscle fatigue [17, 18, 19, 20, 21], muscle pain [22] and for prosthetic control [23, 24, 25].

A flexible piezoelectric thin film sensor has been proposed in several studies for detection of muscle movements [2, 26, 27]. This sensor detect changes in the circumferential length of the skin during muscle contraction [27]. Its flexibility, high sensitivity and the low cost are major advantages for wearable devices that can be used for monitoring skeletal muscle performance and fatigue. The objective of this study was, therefore, to further investigate its ability to assess skeletal muscle performance and fatigue under different dynamics motion.

2. Methods

Simultaneous measurements of surface EMG and piezoelectric (Piezo) signals during different tasks were taken from three healthy subjects mean age 29 $\pm$ 0.8 years.

2.1 Experimental setup and protocol

The experimental system was designed for noninvasive measurements of muscle activity using EMG amplifiers (BIPAC-EMG100C), disposable EL503 vinyl surface electrodes and DT piezoelectric sensors. The signals were sampled at 1 kHz and were analyzed using MATLAB software. Two surface electrodes and two piezoelectric thin film sensors were attached to the same measured muscle. The EMG and Piezo signals were measured: from the biceps during slow and fast elbow flexion with and without strong grip, during different mass lifting (3, 5 and 12.5 kg), during mass lifting of 1 kg for 150 sec and from the gastrocnemius muscle during marching at speeds of 4 and of 10 kph.

2.2 Data analysis

The raw EMG signals were fully rectified to yield the absolute value of the surface EMG signal and filtered with a low pass filter (e.g., Butterworth with a cut-off of 10 Hz) to yield the signal envelopes (IEMG). The root mean square (RMS) of the EMG signal (EMG_RMS) were computed for successive segments of 1 second. For each muscle, the instantaneous mean of EMG_RMS values was calculated over 8 seconds intervals (Fig. 1a–c).

Figure 1.

An example of (a) raw EMG, (b) the EMG_RMS values of each second, (c) the instantaneous mean of EMG_RMS over 8 seconds, (d) raw Piezo and (e) the negative Piezo peaks during biceps flexion.

In order to assess muscle fatigue during mass lifting of 1 kg for 150 sec, the EMG_RMS of EMG signal was divided into four equal time segments. The EMG_RMS values were normalized with respect to the maximal value (EMG_RMSn) and then the means $\pm$ SE of the EMG_RMSn of each segment was calculated. Muscle fatigue was characterized by an increase in the EMG_RMSn with the time segment [28, 29, 30].

The negative values of the raw Piezo signal correlates with muscle contraction. Therefore, only the negative peaks were derived (Fig. 1d and e). The Piezo signal, during the fatigue assessment was divided, as the EMG signal, into the same four equal time segments. The Piezo signal peaks were normalized with respect to their maximal value (Piezon) and then the means $\pm$ SE of the Piezon of each segment was calculated.

Analysis of variance (One way ANOVA) and Tukey honest significant difference were applied to compare the distributions of the average EMG_RMSn and the average Piezon across the time segments and to compare the IEMG amplitudes and the Piezo peaks between the different maneuvers. Value of alpha set at 0.05 and P-values less than 0.05 ( $p<$ 0.05) was used for statistical significance. In addition, regression modeling was done to predict the correlation between the average maximal EMG_RMS and average Piezo peaks of biceps muscles during flexion with 4 different mass lifting and the correlation between the average EMG_RMSn and Piezon with the increase in the time segments.

3. Results

3.1 Increased intensity and load

Increasing contraction intensity during slow and fast elbow flexion increase significantly ( $p<$ 0.001) both IEMG amplitudes and Piezo peaks. Average maximal IEMG (IEMG ${}_{max}$ ) amplitudes and Piezo peaks increases from 132.5 $\pm$ 24 $\mu$ V, 1.13 $\pm$ 0.04 V to 861 $\pm$ 50 $\mu$ V, 1.5 $\pm$ 0.04 V during slow flexion, without and with grip, respectively, and from 116 $\pm$ 10 $\mu$ V, 1.38 $\pm$ 0.05 V to 666 $\pm$ 39 $\mu$ V, 2.32 $\pm$ 0.1 V during fast flexion, without and with grip, respectively. Increasing contractions velocity, both without and with grip, increase significantly the Piezo signal peaks ( $p<$ 0.001). Increasing treadmill marching speed from 4 kph to 10 kph increase significantly ( $p<$ 0.001) both the IEMG amplitudes (average maximal values increased from 30 $\pm$ 1.16 $\mu$ V to 99 $\pm$ 3.7 $\mu$ V) and the Piezo peaks (average peaks increase from 0.5 $\pm$ 0.016 V to 1.1 $\pm$ 0.04 V) that was measured from the gastrocnemius muscle.

Figure 2.

IEMG envelope (a–d) and Piezo signals (e–h) measured from the biceps during elbow flexion while lifting mass of 0 kg (a&e), 3 kg (b&f), 5 kg (c&g) and 12.5 kg (d&h). The correlation between the average Piezo peaks and the average IEMG ${}_{max}$ with the increased mass (i).

During different mass lifting (0 kg, 3 kg, 5 kg, and 12.5 kg), in all 3 subjects, the IEMG signal increased while the Piezo peaks decreased with the increased mass (Fig. 2). The average IEMG ${}_{max}$ increased from 353 $\pm$ 31 $\mu$ V to 912 $\pm$ 51 $\mu$ V while the average Piezo peaks decreased from 1.13 $\pm$ 0.02 V to 0.18 $\pm$ 0.02 V during elbow flexion without and with mass lifting of 12.5 kg. Opposite linear relation ( $R^{2}=$ 0.98) was found between the average IEMG ${}_{max}$ and the average Piezo peaks with the increased mass lifting (Fig. 2i).

3.2 Muscle fatigue

Significant increase in the average EMG_RMSn and significant decrease in the average Piezon of all subjects was found between time segment 1 or 2 to time segment 3 or 4 and between time segments 3 to 4 (Fig. 3). Average values of the EMG_RMSn and Piezon of all subjects increased from 0.54 $\pm$ 0.03, 0.75 $\pm$ 0.02 at time segment 1 to 0.71 $\pm$ 0.02, 0.64 $\pm$ 0.01 at time segment 4 (Fig. 3c and d). Opposite linear relation ( $R^{2}=$ 0.9) was seen between the Piezon and EMG_RMSn (Fig. 3e).

Figure 3.

The average EMG_RMSn (a,c) and the average Piezon (b,d) at each time segment of one subject (a–b) and the average of all subjects (c–d). The correlation between the average EMG_RMSn and the average Piezon with the increase in the time segments (e).

4. Discussion

As was expected, correlation was found between the onset and the duration of the biceps EMG and Piezo signals. Other studies found that the time of Piezo peaks (both positive and negative) provide a good indicator of the starts and ends of contractions [2, 26].

Increasing the biceps contractions velocity increase significantly ( $p<$ 0.001) the Piezo peaks. Increasing the contractions velocity probably caused large vibratory motions that was detected by the sensor and increased the signal peaks. The obtained velocity related increase was also found in other studies measured muscle performance during concentric isokinetic contraction using piezoelectric crystal contact sensor [31, 32].

Elbow flexion with grip increase significantly ( $p<$ 0.001) the IEMG amplitude and Piezo peaks compare to without grip. The low level of effort needed during grip caused an increase in the recruitment the muscle fibers producing morphological changes that were measured by the Piezo signal. This behavior was reported in studies aimed to examine the MMG-RMS at low to high intensity [18, 33] during isometric contraction. To examine the influence of higher level of efforts EMG and Piezo signals were measured during elbow flexion while holding different mass weight. While the IEMG amplitude increased with the mass the Piezo peaks decreased and an opposite linear relation ( $R^{2}=$ 0.9752) was found. Increasing the mass required an increase in motor units recruitments and their firing frequency, that is directly reflected in the measured EMG signal [34, 35]. The Piezo signal, on the other hand, reflects morphological changes during contraction. Holding a mass caused immediate morphological changes that remained throughout the maneuver. The higher the initially morphological change was the smaller morphological changes could be produced. This may explain the reduction seen here in the Piezo peaks.

A significant increase with time in the average EMG_RMSn of the biceps muscle was seen in all subjects during the fatigue maneuver as was expected [34, 35, 36]. The average Piezon, on the contrary, decreased significantly in all subjects, and an opposite linear relation ( $R^{2}=$ 0.9) between the EMG_RMSn and Piezon was found. The increase in the recruitment of more motor units probably reduced the ability to produce morphological changes during the contractions resulting with a decrease in the Piezon. The linear opposite relation found here, therefore, implies that piezoelectric thin film sensor can be used to track muscle fatigue during isokinetic contractions. A decrease in the MMG amplitude measured from the biceps muscle with time during isometric elbow flexion was also explained by minimal dimensional changes during fatigue [33, 38]. Other studies that investigated muscular fatigue through MMG measurements [17, 19, 21, 36, 37], found similar behavior of MMG and EMG signal. These studies, however, used either accelerometer or microphone sensors that detected the vibrations of the muscle fibers during contraction [8]. An increase in the fibers recruitment during muscular fatigue increase in turn the detected vibrations in term of acceleration. Higher acceleration may occur during minimal displacement and thus can correlate to our findings.

5. Conclusion

Increasing contraction intensity and velocity increase significantly both IEMG and Piezo peaks. Opposite linear relation was found between the average IEMG ${}_{max}$ amplitudes and the average Piezo peaks with increasing weight. The significant increase in the average EMG_RMSn and decrease in the average Piezon of all subjects during elbow flexion while holding a mass with time suggesting the ability of this sensor to track muscle fatigue during isokinetic contractions.

Footnotes

Conflict of interest

None to report.

References

Basmajian

De Luca

. Muscles alive: Their functions revealed by electromyography . Williams & Wilkins; 1985.

Fukuda

Ueno

Inoue

. A preliminary study on detection of muscle activity using a flexible AlN piezoelectric thin film sensor. In: Robotics and Biomimetics (ROBIO), 2009 IEEE International Conference on 2009 Dec 19 (pp. 944-948). IEEE.

Chen

Zheng

Guo

Zhu

Chan

Zhang

. Sonomyographic responses during voluntary isometric ramp contraction of the human rectus femoris muscle. European Journal of Applied Physiology. 2012; 112(7): 2603-14.

Duchêne

Goubel

. Surface Electromyogram during Voluntary Contraction: Processing Tools and Relation to Physiological. Critical Reviews in Biomedical Engineering. 1993; 21(4): 313-97.

Tao

Liu

Zheng

Feng

. Gait analysis using wearable sensors. Sensors. 2012; 12(2): 2255-83.

Nadiv

Vachbroit

Gefen

Elad

Zaretsky

Moran

Halpern

Ratnovsky

. Evaluation of fatigue of respiratory and lower limb muscles during prolonged aerobic exercise. Journal of Applied Biomechanics. 2012; 28(2): 139-47.

Hodges

Pengel

Herbert

Gandevia

. Measurement of muscle contraction with ultrasound imaging. Muscle & Nerve. 2003; 27(6): 682-92.

Ibitoye

Hamzaid

Zuniga

Wahab

. Mechanomyography and muscle function assessment: A review of current state and prospects. Clinical Biomechanics. 2014; 29(6): 691-704.

Al-Mulla

Sepulveda

Colley

. A review of non-invasive techniques to detect and predict localised muscle fatigue. Sensors. 2011; 11(4): 3545-94.

10.

Zhou

Ivanov

Zheng

. Estimation and visualization of longitudinal muscle motion using ultrasonography: a feasibility study. Ultrasonics. 2014; 54(3): 779-88.

11.

Watakabe

Mita

Akataki

Itoh

. Mechanical behaviour of condenser microphone in mechanomyography. Medical and Biological Engineering and Computing. 2001; 39(2): 195-201.

12.

Orizio

Esposito

Paganotti

Marino

Rossi

Veicsteinas

. Electrically-elicited surface mechanomyogram in myotonic dystrophy. The Italian Journal of Neurological Sciences. 1997; 18(4): 185-90.

13.

Tong

. The mechanomyography of persons after stroke during isometric voluntary contractions. Journal of Electromyography and Kinesiology. 2007; 17(4): 473-83.

14.

Krueger

Scheeren

Nogueira-Neto

da SN Button

Nohama

. A new approach to assess the spasticity in hamstrings muscles using mechanomyography antagonist muscular group. InEngineering in Medicine and Biology Society (EMBC), 2012 Annual International Conference of the IEEE 2012 Aug 28 (pp. 2060-2063). IEEE.

15.

Marusiak

Jaskólska

Kisiel-Sajewicz

Yue

Jaskólski

. EMG and MMG activities of agonist and antagonist muscles in Parkinson’s disease patients during absolute submaximal load holding. Journal of Electromyography and Kinesiology. 2009; 19(5): 903-14.

16.

Madeleine

Bajaj

Søgaard

Arendt-Nielsen

. Mechanomyography and electromyography force relationships during concentric, isometric and eccentric contractions. Journal of Electromyography and Kinesiology. 2001; 11(2): 113-21.

17.

Tarata

. Mechanomyography versus electromyography, in monitoring the muscular fatigue. BioMedical Engineering OnLine. 2003; 2(1): 3.

18.

Orizio

Gobbo

Diemont

Esposito

Veicsteinas

. The surface mechanomyogram as a tool to describe the influence of fatigue on biceps brachii motor unit activation strategy. Historical basis and novel evidence. European Journal of Applied Physiology. 2003; 90(3-4): 326-36.

19.

Zuniga

Housh

Camic

Hendrix

Schmidt

Mielke

Johnson

. A mechanomyographic fatigue threshold test for cycling. International Journal of Sports Medicine. 2010; 31(9): 636-43.

20.

Stokes

Dalton

. Acoustic myography for investigating human skeletal muscle fatigue. Journal of Applied Physiology. 1991; 71(4): 1422-6.

21.

Kimura

Fujibayashi

Tanaka

Moritani

. Mechanomyographic responses in quadriceps muscles during fatigue by continuous cycle exercise. European Journal of Applied Physiology. 2008; 104(4): 651-6.

22.

Madeleine

Arendt-Nielsen

. Experimental muscle pain increases mechanomyographic signal activity during sub-maximal isometric contractions. Journal of Electromyography and Kinesiology. 2005; 15(1): 27-36.

23.

Zeng

Yang

Cao

Xia

. Hand-motion patterns recognition based on mechanomyographic signal analysis. In: BioMedical Information Engineering, 2009. FBIE 2009. International Conference on Future 2009 Dec 13 (pp. 21-24). IEEE.

24.

Xie

Zheng

Guo

. Classification of the mechanomyogram signal using a wavelet packet transform and singular value decomposition for multifunction prosthesis control. Physiological Measurement. 2009; 30(5): 441.

25.

Alves

Chau

. Automatic detection of muscle activity from mechanomyogram signals: a comparison of amplitude and wavelet-based methods. Physiological Measurement. 2010; 31(4): 461.

26.

Tsukamoto

Ueno

Shima

Tsuji

. Measuring muscle movements for human interfaces using a flexible piezoelectric thin film sensor. In: Engineering in Medicine and Biology Society, 2008. EMBS 2008. 30th Annual International Conference of the IEEE 2008 Aug 20 (pp. 112-116). IEEE.

27.

Han

Kim

. Novel muscle activation sensors for estimating of upper limb motion intention. In: Engineering in Medicine and Biology Society, 2009. EMBC 2009. Annual International Conference of the IEEE 2009 Sep 3 (pp. 3767-3770). IEEE.

28.

Krogh-Lund

Jørgensen

. Myo-electric fatigue manifestations revisited: power spectrum, conduction velocity, and amplitude of human elbow flexor muscles during isolated and repetitive endurance contractions at 30% maximal voluntary contraction. European Journal of Applied Physiology and Occupational Physiology. 1993; 66(2): 161-73.

29.

Ohashi

. Changes in Relations between Surface Electromyogram and Fatigue Level by Repeatlng Fatlgulng Statlc Contractlons. The Annals of Physiological Anthropology. 1993; 12(5): 285-96.

30.

Perlovitch

Gefen

Elad

Ratnovsky

Kramer

Halpern

. Inspiratory muscles experience fatigue faster than the calf muscles during treadmill marching. Respiratory Physiology & Neurobiology. 2007; 156(1): 61-8.

31.

Evetovich

Housh

Stout

Johnson

Smith

Ebersole

. Mechanomyographic responses to concentric isokinetic muscle contractions. European Journal of Applied Physiology and Occupational Physiology. 1997; 75(2): 166-9.

32.

Cramer

Housh

Weir

Johnson

Berning

Perry

Bull

. Gender, muscle, and velocity comparisons of mechanomyographic and electromyographic responses during isokinetic muscle actions. Scandinavian Journal of Medicine & Science in Sports. 2004; 14(2): 116-27.

33.

Orizio

Perini

Veicsteinas

. Muscular sound and force relationship during isometric contraction in man. European Journal of Applied Physiology and Occupational Physiology. 1989; 58(5): 528-33.

34.

Konrad

. The abc of emg. A Practical Introduction to Kinesiological Electromyography. 2005; 1: 30-5.

35.

De Luca

. The use of surface electromyography in biomechanics. Journal of Applied Biomechanics. 1997; 13(2): 135-63.

36.

Madeleine

Farina

Merletti

Arendt-Nielsen

. Upper trapezius muscle mechanomyographic and electromyographic activity in humans during low force fatiguing and non-fatiguing contractions. European Journal of Applied Physiology. 2002; 87(4): 327-36.

37.

Beck

Housh

Johnson

Weir

Cramer

Coburn

Malek

. Comparison of a piezoelectric contact sensor and an accelerometer for examining mechanomyographic amplitude and mean power frequency versus torque relationships during isokinetic and isometric muscle actions of the biceps brachii. Journal of Electromyography and Kinesiology. 2006; 16(4): 324-35.

38.

Kimura

Hamada

Watanabe

Maeda

Oya

Moritani

. Mechanomyographic responses in human biceps brachii and soleus during sustained isometric contraction. European Journal of Applied Physiology. 2004; 92(4-5): 533-9.