Table 4 Assessment of muscle fatigue using MMG and ES

Study 1: examination of prolonged low-force fatigue due to prolonged low- and high-frequency ES

[47]

MMG: uniaxial accelerometer (Bang & Olufsen Technology, Denmark, diameter = 17.6 mm, weight = 2.9 g, 20 pC/ms⁻², 0.1 to 800 Hz); EMG: bipolar Ag/AgCl (Blue Sensor N-00-S, Medicotest, Denmark, diameter = 6 mm); Force: Alpha Beam 250 N (BLH Electronics, USA)

ECR

3 healthy and 4 healthy male subjects, age 27–54 years, height 1.52–1.84 m, weight 56–92 kg, body mass index: 20.9–27.2 kg/m²

Arm flexed at 90º, static wrist extension for 10 min at 10% MVC; over 150 min, the subject was administered a 10-s train at 1 Hz, two 2.5-s trains at 20 Hz, and two 2-s trains at 100 Hz s, with a 30-s rest between each train

The ES was adjusted to obtain 30% of the MVC at 100 Hz, each pulse lasted 0.7 s

LFF showed electromechanical efficiency in the low-force and control experiment for up to 9000 s, as reflected in MMG more than the EMG

Remark: a low-force test is recommended for fatigue development in muscle and recovery prior to low-force exertion

Study 2: analysis of the changes in electromechanical delay components of skeletal muscle exposed to fatigue

[49]

EMG: model ELSCH004 (OT Bioelettronica, Turin, Italy); MMG: accelerometer (model ADXL103, Analogue Devices, Norwood, MA, USA; weight < 1.0 g, sensitivity = 1000 mV/g, range =  ± 1.7 g); force: load cell (model SM-1000 N, Interface, Crowthorne, UK, linear operation between 0 and 1000 N)

GM

20 healthy subjects, age (means ± standard deviations) 23.1 ± 4.2 years, body mass 74.3 ± 11.2 kg, stature 1.77 ± 0.08 m

12 blocks of 10-s stimulation at 35 Hz, pulse duration of 340 us, a duty cycle of 9 s on/1 s off for a duration of 120 s was delivered after 2-Hz, a 5-mA increasing amplitude was administered for motor unit stimulation

All of the delays lengthened the contraction based on different onsets and kinetics. The changes in the cross-bridge and muscle tendon unit (MTU) mechanical properties occurred later compared with the electrochemical events

Remark: the delay in the lengthening of mechanical events suggests that these were the most affected by fatigue

Future work: further studies are needed to evaluate the electrochemical and electromechanical alternations at the motor tendon units and muscle junctions under fatigue conditions

Study 3: evaluation of the inter- and intra-operator reliability of the measurements and the effects of fatigue on different \({\mathrm{Delay}}_{\mathrm{TOT}}\) components

[50]

MMG: accelerometer (model ADXL103, Analogue Devices, Norwood, MA, USA; weight =  < 1.0 g; sensitivity = 1,000 mV/g; range =  ± 1.7 g); force: load cell (mod. SM-1000 N, Interface, UK); EMG: linear array of four electrodes (model ELSCH004, OT Bioelettronica, Turin, Italy)

GM

16 healthy male subjects, age (means ± SDs) 25.0 ± 3.9 years, body mass 77.5 ± 13.8 kg, stature 1.79 ± 0.08 m

A biphasic 50-Hz pulse with a 340-µs duration at 110% of the M-wave amplitude was administered during a 3-s tetanic stimulation with a 5-min rest before a fatiguing tetanic pulse of 120 s at 35 Hz was delivered to the GM. The pre-fatiguing pulse was repeated at 1, 2, and 7 min

The ICC was 0.874–0.996, and the SEM was 0.78 and 6.61% before fatigue. The reliability was 0.781 to 0.981, and the SEM was 1.78 to 8.71%. All the variables were reliable within an inter-parameter operability of 0.847 to 0.999 after fatigue

Remark: the intra- and inter-operator reliability of individual \({Delay}_{TOT}\) components increases, and the results provide a valid indication for the monitoring of physiological and pathological changes

Study 4: effects of fatigue on \({\mathrm{Delay}}_{\mathrm{TOT}}\) components and their inter-session and inter-day reliability

[51]

MMG: accelerometer (model ADXL103, Analogue Devices, Norwood, MA, USA; device weight =  < 1.0 g; sensitivity = 1000 mV/g; range =  ± 1.7 g); EMG (model ELSCH004, OT Bioelettronica, Turin, Italy)

GM

17 healthy male subjects, age 24.3 ± 3.4 years, body mass 77.8 ± 14.3 kg, stature 1.79 ± 0.08 m

A biphasic 50-Hz pulse with a 340-µs duration at 110% of the M-wave amplitude was administered during a 3-s duration tetanic stimulation with a 10-min rest before a fatiguing tetanic pulse of 120 s was delivered at 35 Hz to the GM

\({\mathbf{D}\mathbf{e}\mathbf{l}\mathbf{a}\mathbf{y}}_{\mathbf{T}\mathbf{O}\mathbf{T}}\), R-∆t F-MMG_p-p, and \({\mathbf{D}\mathbf{T}}_{\mathbf{S}\mathbf{l}\mathbf{o}\mathbf{w}}\) were positively correlated. Fatigue changed the duration of the experiment and the start of force decays but had no effect on its duration

Remarks:

1. Fatigue might prolong R-EMD components, which might indicate physiological recovery after physical or rehabilitation exercises

2. After fatigue, the constant elongation of the R-∆t MMGR-F without changes in R-∆t F-MMGPP might be due to increases in the spatial relationship between blood vessels and muscle fibers

Future work: the reoccupation of the squeezed out interstitial fluid might be attributed to alterations in the return of the MMG signal to baseline. Further studies are needed

Study 5: evaluation of the reliability of MMG for determining the evoked changes during muscle relaxation

[52]

MMG: (model 10ADXL103, Analogue Devices, Norwood, MA, USA; weight =  < 1.0 g; 1000 mV/g; range =  ± 1.7 g); EMG: (model ELSCH004, OT Bioelettronica, Turin, Italy); Force: (model SM-1000 N, Crowthorne, UK)

GM

23 HV male subjects, age 25.1 ± 3.8 years, body mass 78.2 ± 15.3 kg, stature 1.81 ± 0.05 m

A biphasic 50-Hz pulse for 340 µs at 110% of the M-wave amplitude was administered during a 3-s tetanic stimulation with a 10-min rest before a fatiguing tetanic pulse for 120 s at 35 Hz was delivered to the GM

High ICC values between the MMG and force parameters was obtained at different experimental sessions on different days

Remark: the correlation between \({\mathrm{R}-\mathrm{MMG}}_{\mathrm{PP}}\) and force confirm the effect of mechanisms of muscle fatigue that modify the extent of velocity and force relaxation

Future work: full relaxation of the muscle under voluntary dynamic and isometric contractions should be investigated

Study 6: analysis of the use of TMG and PMT to evaluate peripheral fatigue-induced alterations in mechanical and contractile properties

[53]

ES: (5 cm², Axelgaard, USA); Contractile properties detection: TMG (BMC Ltd., Ljubljana, Slovenia)

GM

21 HV male subjects, age (means ± SDs) 21.3 ± 3.4 years, height 182.0 ± 6.1 cm, mass 79.5 ± 10.0 kg

1-ms pulse; amplitude of 20 mA, 10-mA increase to evoke Dm. An inter-pulse rest of 10 s was used to lower both fatigue and potentiation. The fatigue protocol was 15 pulses (1 every 100 ms) at approximately 110 mA over 5 min

PMT from the plantar flexor increased, and the TMG Dm decreased, which confirmed muscle stiffness after fatigue

Remark: a decrease in the TMG Dm is not associated with Vc, but the results are limited to healthy individuals and cannot be applied to non-superficial muscles

Future work: future verification of fatigue changes in various cohorts should be performed

Study 7: prediction of the muscle fatigue in SCI patients using SVM

[54]

ES: RehaTrode (HASOMED, Germany); MMG: accelerometer (Sonostics BPS-IIVMG transducers; 20–200 Hz, 30 V/g, 10 g)

VL, RF, and VM

5 SCI patients classified as Class A and B according to ASIAIS

30-min NMES-cycling at 120 mA, 30 Hz, biphasic, pulse width of 400 ± 400 μs. The MFCC and RMS MMGs were trained and tested based on SVM

Contractions correctly identified as non-fatigued and fatigued had higher MFCC compared with RMS values

Remark: both MFCC and RMS showed that fatigued muscle contractions overlapped with non-fatigued muscle contractions, which resulted in insufficient prediction

Future works:

1. Multiple experimental trials and analyses of the effects of the window size of MMG signals should be performed to improve the accuracy of the SVM classifier

2. The nature of MMG signals that influence physiological properties and the physical environment should be investigated

Study 8: assessment of fatigue based on MMG and torque responses

[55]

ES: two 9-cm × 15-cm self-adhesive electrode (RehaTrode, HASOMED, Germany); MMG: (Sonostics VMG BPS II Transducer, frequency = 20–200 Hz, sensitivity = 50 V/g)

VL, VM and RF

5 male and 1 female subjects with complete chronic SCI

30 Hz, 400-μs pulse width, 90–120 mA to the quadriceps; 30 Hz, 60–120 mA, 300 μs was used to elicit NMES-leg cycling exercise from the quadriceps; and 58–90 mA was applied to the hamstrings for 30 min

The MMGmpf, MMGrms and NMES-cycle time altered similarly to the epT during pre- and post- fatigue in a 2-day test

Remark: the high fatigability of RF might lead to limited cycling exercises in SCI subjects

Future work: the reliability of muscle fatigue assessment during functional recovery using NMES-cycling in SCI patients should be further investigated

Study 9: quantification of fatigue under repeated functional electrical excitation

[56]

MMG measurement device: LIS311, STMicroelectronics, USA

TA

21 healthy male subjects, age 24.0 ± 2.0 years, height 174.1 ± 6.9 cm, weight 75.3 ± 13.2 kg

60 Hz, 240 μs, relaxation times of 0.5 s, 2.0 s, 0.5 s, and 1.0 s. MMG signals were collected for 30 min. During MVC and ES, the torque was measured offline using a dynamometer

During muscle fatigue, the MMGpp, convex-hull volume, and convex-hull area linearly decreased with decrease in the mean and median frequencies

Remark: the Lempel–Ziv symbolization technique exhibited the best performance for complex MMG feature reduction

Future works:

1. Based on Lempel–Ziv symbolization, further studies should address the required threshold for individuals under NMES

2. Repeated evoked fatigue was observed under isometric conditions, and isokinetic and/or isotonic conditions should be investigated

Study 10: analysis of the mechanical fatiguing phenomenon that develops during ES in sport training or rehabilitation protocols

[57]

MMG: ADXL202JE (Analogue Devices, Inc., Norwood, MA, USA); EMG: silver bar (diameter = 1 mm, length = 5 mm, inter-electrode = 10 mm); force: load cell (Interface, model SM-100 N, operation range = 0 and 100 N)

BB and VL

10 healthy subjects, age 20–50 years

6 potentiation pulses of 100 Hz with a rest of 1 s followed by a fatigue protocol of 50 Hz for 2 s and 2 Hz for 25 s. Normalized MMG and pT were linearly used for fatigue evaluation

The torque and MMG decreased linearly from 100% of their initial values to 50 and 60% for the BB and to 43 and 47% for the VL, respectively

Remark: accurate screening of muscle mechanical fatigue should eradicate muscle tendon unit insertion and joints