Vromans and Faghri: Electrical Stimulation Frequency and Skeletal Muscle Characteristics: Effects on Force and Fatigue

Electrical Stimulation Frequency and Skeletal Muscle Characteristics: Effects on Force and Fatigue

Abstract

This investigation aimed to determine the force and muscle surface electromyography (EMG) responses to different frequencies of electrical stimulation (ES) in two groups of muscles with different size and fiber composition (fast- and slow-twitch fiber proportions) during a fatigue-inducing protocol. Progression towards fatigue was evaluated in the abductor pollicis brevis (APB) and vastus lateralis (VL) when activated by ES at three frequencies (10, 35, and 50Hz). Ten healthy adults (mean age: 23.2 ± 3.0 years) were recruited; participants signed an IRB approved consent form prior to participation. Protocols were developed to 1) identify initial ES current intensity required to generate the 25% maximal voluntary contraction (MVC) at each ES frequency and 2) evaluate changes in force and EMG activity during ES-induced contraction at each frequency while progressing towards fatigue. For both muscles, stimulation at 10Hz required higher current intensity of ES to generate the initial force. There was a significant decline in force in response to ES-induced fatigue for all frequencies and for both muscles (p<0.05). However, the EMG response was not consistent between muscles. During the progression towards fatigue, the APB displayed an initial drop in force followed by an increase in EMG activity and the VL displayed a decrease in EMG activity for all frequencies. Overall, it appeared that there were some significant interactions between muscle size and fiber composition during progression towards fatigue for different ES frequencies. It could be postulated that muscle characteristics (size and fiber composition) should be considered when evaluating progression towards fatigue as EMG and force responses are not consistent between muscles.




Ethical Publication Statement

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Fatigue is defined as drop in force from pre-to -post repeated ES-induced muscle contraction protocols1,2 or declines in ES-induced torque.3,4 However, EMG changes during fatiguing ES protocols are not consistent and have shown both increases or decreases in electrical activity, defined by root-mean-square (RMS) amplitude, peak-to-peak amplitude and area under the signal.3,5 Yet, with frequency assessment, research more consistently observed decreases in mean frequency (MNF) and median frequency (MDF) as fatigue progressed.6,7 Furthermore, effectively monitoring EMG activity of the muscle throughout the duration of ES also creates added complexity. This is due to ES-induced contractions cluttering the pure EMG signals, masking the unique muscle activity of the signals. These have been attributed to either close proximity of ES electrodes to the EMG detection region or other external noises.5,8 Evidently, these challenges limited researchers’ abilities to establish a clear understanding of muscle responses to ES and fiber recruitment pattern with ES. Many approaches have been proposed for artifact removal including software filters incorporating wavelet transformations,9 hardware devices that employ switch circuits to shut down amplifier control,10 and more recently a new approach called Empirical Mode Decomposition (EMD). EMD implements a repeated sifting process that identifies local maxima and minima in the signal to form envelopes that are extracted as intrinsic mode functions (IMFs) in order to separate muscular components from the artifacts.9,11

This application seems to be more promising and was applied to filtering EMG signals in the present study. Moreover, muscles of different size and fiber composition may respond differently to distinctive ES frequency and current intensity and such characteristics may also play a role in the development of fatigue. Understanding these characteristics may help in the development of more effective ES protocols that delay fatigue while maximizing muscle response during rehabilitation treatment. For example, research has suggested that larger muscles, like the predominantly type II vastus lateralis,12 generally favor the recruitment of additional motor units as a means of increasing force production until the force demand ranges from 60-90% MVC; thereafter, these muscles rely on greater firing rates to achieve additional force. On the contrary, smaller muscles, such as the predominantly type I abductor pollicis brevis (APB),12 utilize the recruitment of additional motor units until force production nears ~30%MVC, after which increases in firing rate enhance force production.13 Related response variability may be true during ES stimulation when selecting the frequency of ES firing for the purpose of either force generation or improving muscle integrity and rehabilitation training. Several studies have examined how manipulation of stimulation frequency, current intensity, and pulse-width parameters impact ES-induced fatigue,14-18 but to our knowledge little research has examined the development of fatigue with consideration of combined effects of muscle sizes with fiber composition variability in the same group of subjects. Additionally, more recent interest has directed ES application towards tissue regeneration where research has shown that ES aids cell proliferation in connective tissue, improves the rate of new collagen formation in injured tendons,19 and reverses long-term denervation muscle atrophy.20 Therefore, it is important to understand how ES affects different types of fibers in order to tailor the stimulation to specific muscle types to enhance tissue growth as well as healing while promoting functional movements. The purpose of this investigation was to evaluate the effect of different ES frequencies (10, 35, and 50Hz) on the development of muscle fatigue in muscles of different size and fiber composition (APB and VL) utilizing ES current intensity, force, and EMG activity as outcome measures. We hypothesized that fatigue response to ES-induced contractions will be different between muscles and will depend on ES characteristics. Accordingly, muscle size and fiber composition should be carefully planned when developing ES-induced muscle contraction protocols.

Materials, Methods

Participants

Ten healthy male and female participants (23.2 ± 3.0 years) with no known history of musculoskeletal or cardiovascular problems, or known allergy to surface electrodes or adhesive tape were recruited. Full-board IRB approval and written consent was obtained prior to participation.

Instruments

ES was delivered using the Respond Select® (Empi, Inc., St. Paul, Minnesota) neuromuscular electrical stimulation system. Self-adhesive, reusable, latex-free bipolar stimulating surface electrodes (90x50mm square for VL; 3.5cm round for APB) were placed on the skin following Empi, Inc. guidelines (Figure 1 – rightside images) and connected to the ES device for muscle activation. EMG signals were acquired through the Nexus-10 EMG device (MindMedia B.V., Netherlands). Sampling frequency was set at 2048Hz. Pre-gelled, self-adhesive Ag/AgCl EMG surface electrodes were placed across the muscle belly in parallel with the muscle fibers according to Surface ElectroMyoGraphy for the Non-Invasive Assessment of Muscles (SENIAM) guidelines. Isometric forces for the APB were measured using the PasPort high resolution force sensor (range: ± 50N, resolution: 0.002N; PASCO Scientific Inc., Roseville, CA) and for the VL using the Shimpo Instruments (ELECTROMATIC Equipment Co., Cedarhurst, NY) Javelin FGV-HXY Force Gauge (maximal capacity: 500lb/250kg; accuracy: +/-0.2% F.S.).

Experimental Set-up

Protocols were developed for the APB and VL to ensure stabilization and isolation of the related muscles during testing. For APB testing, subjects were seated with the arm in 90° elbow flexion. Hand and forearm muscles were isolated via resting on a board that blocks the fingers, only to let the thumb contract.21 The thumb force sensor was affixed to a rod projecting from the table surface so as to be positioned facing the palm of the hand at thumb height (Figure 1a). For VL testing, subjects were seated in an upright chair (hips at 90° flexion) as such that feet did not make contact with the ground and the back of the knee joint contacted the edge of the chair. A strap was placed firmly around the waist to isolate the quadriceps.21,22 The force sensor was positioned facing the lower anterior leg, and was affixed from the side of the chair (Figure 1b).

Procedure

The order at which the frequencies were delivered and of which muscle was first stimulated was randomized. During an acclimation session, each participant was familiarized with appropriate posturing and execution of MVC. Furthermore, each participant was familiarized with ES sensation to ensure they felt comfortable during the actual procedure.

Part One: Pre-fatigue MVC

For each muscle at each of three stimulation frequencies (10, 35, and 50Hz), all participants performed two pre-fatigue MVC trials (holding contraction for 5 seconds then relaxing) of isometric knee and thumb extension. Muscle force and EMG was simultaneously recorded. Verbal encouragement was offered from the rater. A minimum of 30 seconds of rest was given between the two trials to allow adequate recovery. The MVC values were averaged and recorded as the preMVC value. The force measure that equated to 25%preMVC was then calculated and a contraction to 25%MVC was performed while the EMG was recorded. The EMG signals were then used for comparison with EMG signals obtained during ES-induced contraction to 25%MVC following EMD filtering.

Part Two: ES-induced progression towards fatigue

Prior to the fatiguing protocol for a given ES frequency, each participant’s required stimulation current intensity to achieve 25%preMVC force through ES-induced contraction was determined. This value was then set as a constant for the fatiguing protocol and the timer began to record time to fatigue. Each participant was stimulated at a randomized frequency order (10, 35, and 50Hz) with rectangular, biphasic pulses delivered at pulse-widths of 300us. Pulses were delivered at a sequence of 4sON/4sOFF.14 Stimulation concluded when the muscle fatigued, defined as a 50% drop in force from the initial 25%preMVC force for three consecutive contractions,14 or at the expiration of 30 minutes of ES. EMG was recorded throughout the protocol. The timer stopped once fatigue or 30 minutes was reached. At least 30 minutes of rest was given in-between fatiguing protocols for each muscle.

Signal Analysis

All EMG signals were initially filtered using a Butterworth filter (20-500Hz cutoff, 5th-order) inherent in the device’s BioTrace+ software (MindMedia B.V., Netherlands). Data was then transferred to MATLAB (The MathWorks, Inc., Natick, Massachusetts) for further processing. The middle two seconds of ES-induced EMG data was extracted at 11 time points, denoted as % of total time to fatigue (TTF), for each participant (initial, 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100% of TTF). EMD was applied to all ES-induced signal data. EMD was previously described in Pilkar et al.19 RMS averaging was performed on all ES signals (post-EMD filtered). EMD-filtered ES signals were also full-wave rectified and noise minimized with a 4th-order, zero-lag, low-pass filter at 5Hz cutoff frequency to form a linear envelope. From the linear envelope, integrated EMG (iEMG) was calculated. Additionally, the Fast Fourier Transform was applied to the full-wave rectified signal and the MDF was calculated.

Statistical Analysis

After testing for normality, ES current intensities (mA) required to reach 25%MVC and changes in ES-induced force and EMG parameters at each time point were evaluated for statistical significance using ANOVAs and Tukey post-hoc tests. Values are reported as mean ± standard error. Significance was defined by a p-value set at p<0.05. SAS V9.4 was used for all statistical analysis.

Results

ES Current Intensity

Generally, the APB required lower initial ES current intensity to achieve the required force (25%MVC) compared to the VL (mean APB: 15.77±1.15mA, mean VL: 46.77±1.15mA, p<0.0001) for all frequencies. The required current intensity to achieve force at 10Hz was significantly greater (p<0.0001) than at 35Hz or 50Hz for the APB (mean at 10Hz: 20.9±0.97mA, mean at 35Hz: 13.3± 1.14mA, mean at 50Hz: 13.1± 0.96mA) and non-significantly greater for the VL (mean at 10Hz: 51.4±2.42mA, mean at 35Hz: 44.9±2.85mA, mean at 50Hz: 44.0±2.60mA). The muscle type*frequency interaction effect significantly influenced stimulation intensity (p<0.0001) in all cases. For example, APB at 10Hz differed significantly from VL at 10, 35, and 50Hz as did APB at 35Hz and APB at 50Hz (Figure 2).

ES-induced Fatigue

Overall, when combining all frequencies for both muscles, force significantly declined with progression of ES-induced contraction (p<0.0001), indicating muscle fatigue. Comparisons between frequencies and progression of fatigue for each muscle indicated that ES of the APB at 10, 35, and 50Hz lead to significant force decline after 40%, 10%, and 70%TTF, respectively for each frequency. Corresponding to these points (40%, 10%, and 70%TTF) of significant force decline, a sudden increase in EMG amplitude was observed. Observations in the VL at 10, 35, and 50Hz revealed significant force decline after 60%, 10%, and 80%TTF, respectively for each frequency. Corresponding to these points (60%, 10%, and 80%TTF) of significant force decline, a sudden decrease in EMG amplitude was observed. Figures 3-5 display the respective electrical activity shifts for each muscle during stimulation at each frequency.

Discussion

Electrical stimulation is applied during rehabilitation to promote functional gains through increased force generation and improved muscle integrity. In addition to the reported effects of increased cell proliferation and angiogenesis,19,23 enhanced muscle integrity and force generation is further achieved by the ES-induced exercise-like movement that leads to fiber type transformation and increased muscle cross-sectional area.20,24-31 With this knowledge, the aim of this investigation was to better understand how different muscle fiber compositions can impact optimal ES performance, particularly to delay fatigue development. Research comparing the application of ES on muscles of varying fiber proportion in the same group of subjects is limited. The result of our study supports our hypothesis that the frequency of ES may play a role in force generation and fatigue development of muscles irrespective of size and characteristics. Further, muscles of different size and fiber composition, seem to behave differently as fatigue progresses. The results show that for both the APB and VL there was a trend in requiring higher ES current intensity during lower frequency (10Hz) in order to achieve the initial force (in our study 25%MVC). Furthermore, the initial ES intensity required to achieve 25%MVC was significantly greater for the VL muscle compared to APB. This difference can be explained by the greater proportion of type II fibers, which have higher firing thresholds and larger diameters, in the VL. Kuriki et al. reported that muscle control differs between sizes.32 Small muscles, like the APB, assist in fine motor movements by innervating fewer fibers per motor unit; therefore, they are less force producing in exchange for greater precision. In contrast, large muscles, like the VL, assist in gross motor function, generating significantly greater forces than smaller muscles by innervating upwards of 100 to 1000 fibers per motor unit.33 Therefore, greater intensity is necessary to activate the larger motor units of the large muscle. Moreover, in the VL, it was reported that recruitment thresholds range from ~5.5 to 22.7 pulses per second34 and given that more fast-twitch fibers tend to be superficially located,12 minimal motor unit recruitment will be achieved at low-levels of stimulation frequency compared to 35 and 50Hz, the latter two which surpass the recruitment threshold for many motor units. Since the APB contains lower-threshold motor units, the recruitment threshold range suggests that stimulation delivered at frequencies above 10Hz will greatly maximize recruitment and therefore contribute to intensity being less influenced by frequency. In conclusion, the results of our investigation indicated that an association exists between ES-induced force decline and the relative EMG responses. The corresponding sudden increase in electrical activity of the APB and decrease in electrical activity of the VL when ES-induced force declined may be due to the better glycogen storage maintenance and utilization ability of type I fibers. Gregory et al. evaluated the influence of metabolic characteristics and phenotypic expression of individual fibers in predicting glycogen utilization during ES.34 Outcomes demonstrated that both an enzyme ratio of succinate dehydrogenase and quantitative-actomyosin ATPase (SDH:qATPase) and fiber phenotype significantly predicted glycogen utilization, where type I fibers are the most proficient at utilizing and maintaining glycogen stores.34 Therefore, it is possible that the fibers of the APB are capable of producing a sudden electrical activity increase at the moment of significant force decline as a final mechanism attempting to sustain force before fatigue while the fast-fatiguing type II fibers do not sustain the same energy capacity and thus, the electrical activity drops at the moment of significant force decline. Furthermore, studies have shown that to increase muscle strength the force production should be maintained at or above 60%MVC. Though our research only stimulated to 25%MVC, it appears that higher frequencies such as 50Hz could be used to increase muscle strength. This is suggested because ES at 10Hz, particularly for the small APB, required increases in intensity beyond some participants’ pain thresholds and therefore may not be tolerable, but the higher frequencies were more comfortable and able to recruit more motor units. Thus, lower frequencies may be more applicable to improve muscle integrity without intention to increase force and higher frequencies may be suitable to promote increases in force and muscle strength. The observed relationship between the initial moment of significant force decline and shift in electrical activity should assist clinicians and researchers when developing ES protocols to minimize fatigue based on muscle size and fiber composition. Based on significant differences observed between muscle force decline and ES frequencies at 10, 35 and 50Hz, we propose to utilize higher frequencies for increasing muscle force (for example in athletic training) and lower frequencies due to their slower onset of fatigue for rehabilitation programs in which force is not the main outcome. The findings from this investigation may provide some guidelines for health professionals to adjust the ES current intensity based on fatigue development to encourage the recruitment of additional motor units to sustain force generation and prevent early fatigue.

List of acron yms

APB

abductor pollicis brevis

EMD

empirical mode decomposition

EMG

electromyography

ES

electrical stimulation

IMFs

intrinsic mode functions

IRB

Institutional Review Board

MNF

mean frequency

MVC

maximal voluntary contraction

qATPase

quantitative-actomyosin ATPase

RMS

root-mean-square

SD

succinate dehydrogenase

SENIAM

Non-Invasive Assessment of Muscles

TTF

time to fatigue

VL

vastus lateralis

Acknowledgments

Funding Disclosure: This work is supported in part by the National Science Foundation under EFRI Grant 1332329. We would like to thank Magdalena Wegrzyniak and Sarina Moghadam for their help with data collection and the School of Engineering machine shop for their aid in constructing dynamometer-holding devices.

References

1. 

C Gregory, W Dixon, C. Bickel Impact of varying pulse frequency and duration on muscle torque production and fatigue. Muscle Nerve 2007;35:504-9.

2. 

N Chesler, W. Durfee Surface EMG as a fatigue indicator during FES-induced isometric muscle contractions. J Electromyogr Kinesiol 1997;7:27-37.

3. 

M Papaiordanidou, D Guiraud, A. Varray Does central fatigue exist under low-frequency stimulation of a low fatigue-resistant muscle?. Eur J Appl Phsiol 2010;110:815-23.

4. 

F Alexandre, G Derosiere, M Papaiordanidou. Cortical motor output decreases after neuromuscular fatigue induced by electrical stimulation of the plantar flexor muscles. Acta Physiol 2015;214:124-34.

5. 

E Estigoni, C Fornusek, N Hamzaid. Evoked EMG versus muscle torque during fatiguing functional electrical stimulation-evoked muscle contractions and short-term recovery in individuals with spinal cord injury. Sensors 2014;14:22907-20.

6. 

T Hamada, T Kimura, T. Moritani Selective fatigue of fast motor units after electrically elicited muscle contractions. J Electromyogr Kinesiol 2004;14:531-8.

7. 

K Johnson, S Edwards, C Van Tongeren, P. Bawa Properties of human motor units after prolonged activity at a constant firing rate. Exp Brain Res 2003;154:479-87.

8. 

M Ibitoye, E Estigoni, N Hamzaid. The effectiveness of FES-evoked EMG potentials to assess muscle force and fatigue in individuals with spinal cord injury. Sensors 2014;14:12598-622.

9. 

A Andrade, S Nasuto, P Kyberd. EMG signal filtering based on empirical mode decomposition. Biomed Signal Process Control 2006;1:44-55.

10. 

D Zhang, WT Ang. Reciprocal EMG controlled FES for pathological tremor suppression of forearm. Conf Proc IEEE Eng Med Biol Soc 2007.

11. 

RB Pilkar, M Yarossi, G. Forrest Empirical mode decomposition as a tool to remove the function electrical stimulation artifact from surface electromyograms: preliminary investigation. Conf Proc IEEE Eng Med Biol Soc 2012;1847-50.

12. 

J Polgar, M Johnson, D Weightman, D. Appleton Data on fibre size in thirty-six human muscles. J Neurol Sci 1973;19:307-18.

13. 

R Andrzejewska, A Jaskólski, A Jaskólska. Electromyogram features during linear torque decrement and their changes with fatigue. Eur J Appl Phsiol 2014;114:2105-17.

14. 

M Behringer, S Grützner, J Montag. Effects of stimulation frequency, amplitude, and impulse width on muscle fatigue. Muscle Nerve 2015;53:608-16.

15. 

B Dreibati, C Lavet, A Pinti, G. Poumarat Influence of electrical stimulation frequency on skeletal muscle force and fatigue. Ann Phys Rehabil Med 2010;53:266-77.

16. 

J Wegrzyk, A Fouré, Y Le Fur. Responders to wide-pulse, high-frequency neuromuscular electrical stimulation show reduced metabolic demand: A 31P-MRS study in humans. PloS One 2015;10:e0143972.

17. 

C Bickel, J Slade, L VanHiel. Variable-frequency-train stimulation of skeletal muscle after spinal cord injury. J Rehabil Res Dev Clin Suppl 2004;41:33.

18. 

L Chou, S. Binder-Macleod The effects of stimulation frequency and fatigue on the force–intensity relationship for human skeletal muscle. Clin Neurophysiol 2007;118:1387-96.

19. 

N Maffulli, JP Furia. Rotator cuff disorders: Basic science and clinical medicine. JP Medical Ltd 2012.

20. 

H Kern, S Boncompagni, K Rossini. Long-term denervation in humans causes degeneration of both contractile and excitation-contraction coupling apparatus, which is reversible by functional electrical stimulation (FES): A role for myofiber regeneration? J Neuropathol Exp Neurol 2004;63:919-31.

21. 

K Stratton, P. Faghri Electrically and hybrid-induced muscle activations: effects of muscle size and fiber type. Eur J Transl Myol 2016;26:6163. eCollection 2016 Jun 13.

22. 

R Bohannon, J Kindig, G Sabo. Isomeric knee extension force measured using a handheld dynamometer with and without belt-stabilization. Physiother Theory Pract 2011;28:562-8.

23. 

S Kanno, N Oda, M Abe. Establishment of a simple and practical procedure applicable to therapeutic angiogenesis. Circulation 1999;99:2682-7.

24. 

JL Andersen, T Mohr, F Biering-Sørensen. Myosin heavy chain isoform transformation in single fibres from m. vastus lateralis in spinal cord injured individuals: Effects of long-term functional electrical stimulation (FES). Pflügers Arch 1996;431:513-8.

25. 

R Crameri, A Weston, M Climstein. Effects of electrical stimulation‐induced leg training on skeletal muscle adaptability in spinal cord injury. Scand J Med Sci Sports 2002;12:316-22.

26. 

S Boncompagni, H Kern, K Rossini. Structural differentiation of skeletal muscle fibers in the absence of innervation in humans. Proc Natl Acad Sci USA. 2007;104:19339-44.

27. 

H Kern, U Carraro, N Adami. One year of home-based functional electrical stimulation (FES) in complete lower motor neuron paraplegia: recovery of tetanic contractility drives the structural improvements of denervated muscle. Neurol Res. 2010;32:5-12. DOI:1 0.1189/184313209X385644

28. 

H Kern, U Carraro, N Adami. Home-based functional electrical stimulation rescues permanently denervated muscles in paraplegic patients with complete lower motor neuron lesion. Neurorehabil Neural Repair 2010;24:709-21. Doi 10.1177/1545968310366129. Epub 2010 May 11.

29. 

H Kern, U. Carraro Home-based functional electrical stimulation (h-b FES) for long- term denervated human muscle: history, basics, results and perspectives of the Vienna rehabilitation strategy. Eur J Transl Myol. 2014;24:27-40.

30. 

U Carraro, H Kern, P Gava. Biology of Muscle Atrophy and of its Recovery by FES in Aging and Mobility Impairments: Roots and By-Products. Eur J Transl Myol 2015;25:221-30. doi: 10.4081/ejtm.2015.5272.

31. 

H Kern, C Hofer, S Loefler. Atrophy, ultra-structural disorders, severe atrophy and degeneration of denervated human muscle in SCI and Aging. Implications for their recovery by Functional Electrical Stimulation, updated 2017. Neurol Res 2017;39:660-6. doi: 10.1080/01616412 .2017.1314906. Epub 2017 Apr 13.

32. 

HU Kuriki, FM De Azevedo, LSO Takahashi. The relationship between electromyography and muscle force: EMG methods for evaluating muscle and nerve function. InTech 2012.

33. 

M Trevino, T Herda, A Fry. Influence of the contractile properties of muscle on motor unit firing rates during a moderate-intensity contraction in vivo. J Neurophysiol 2016;116:552-62.

34. 

CM Gregory, RH Williams, K Vandenborne, GA Dudley. Metabolic and phenotypic characteristics of human skeletal muscle fibers as predictors of glycogen utilization during electrical stimulation. Euro J Appl Physiol 2005;95:276-82.

Fig 1.

Experimental setup for the (a) APB and (b) VL.

ejtm-27-4-6816-g001.jpg
Fig 2.

Required stimulation intensity to achieve initial contractile force (25%MVC). (*) denotes significant differences between muscles (p<0.05). (+) denotes significant differences within muscle (p<0.05).

ejtm-27-4-6816-g002.jpg
Fig 3.

Force and electrical activity changes during ES at 10Hz with progression towards fatigue.

ejtm-27-4-6816-g003.jpg
Fig. 4.

Force and electrical activity changes during ES at 35Hz with progression towards fatigue.

ejtm-27-4-6816-g004.jpg
Fig 5.

Force and electrical activity changes during ES at 50Hz with progression towards fatigue.

ejtm-27-4-6816-g005.jpg
Abstract views:
291

Views:
PDF
49
HTML
79

Article Metrics

Metrics Loading ...

Metrics powered by PLOS ALM


Copyright (c) 2017 Maria Vromans, Pouran Faghri

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
 
© PAGEPress 2008-2018     -     PAGEPress is a registered trademark property of PAGEPress srl, Italy.     -     VAT: IT02125780185