Previous studies related to the monitoring of muscle fatigue during dynamic movement have focused on detecting the accumulation of muscle fatigue. However, the accumulation and recovery of muscle fatigue should be detected in dynamic muscle contraction while the muscle load is constantly changing. The purpose of this study is to investigate the development and recovery of muscle fatigue under conditions of dynamic muscle contraction following continuous load changes.
Meanwhile, using the EMG signal from the biceps brachii muscle, fatigue detections were performed based on both dynamic measurements during each dynamic muscle contraction task and isometric measurements during isometric muscle contraction just before and after each task. They were compared to check the validity of the muscle fatigue monitoring method using Wavelet transform with EMG signal from dynamic measurements. It shows that the development and recovery of muscle fatigue were detected in 2 kg and 1 kg intensity tasks, respectively.
It suggests that monitoring muscle fatigue under conditions of dynamic muscle contraction with the wavelet transform was valid for continuously detecting the development and recovery of muscle fatigue. The result also shows the possibility of real-time monitoring of muscle fatigue in industry and may suggest a guideline in the design of a human-robot interaction system based on monitoring the user's muscle fatigue.
INTRODUCTION
Research Background
- Evaluation of muscle fatigue
- Muscle fatigue assessment based on EMG
- Application of EMG based muscle fatigue assessment
Therefore, measuring lactate concentration in a muscle using blood samples is one of the methods to assess muscle fatigue. Cobb and Forbes (1923); Lippold, Redfearn and VuČO (1960) observed that the amplitude of the EMG signal increases as one of the indicators of muscle fatigue during an isometric contraction. Therefore, in general, muscle fatigue is assessed by combining with spectral analysis (Joint EMG Spectrum and Amplitude Analysis, JASA).
In a frequency domain, the shift of the EMG signal spectrum towards lower frequencies can be detected during sustained muscle contraction as a dominant indicator of the development of muscle fatigue. Therefore, how to assess changes in the frequency content of EMG signals resulting from muscle fatigue has been investigated over the last decade. However, the application of the muscle fatigue recovery detection method has rarely been investigated.
However, these were limited to assessing recovery from muscle fatigue during rest conditions. The detection of muscle fatigue before robotic intervention is also important to properly monitor follow-up assistance.
Research Objectives
Research hypotheses
METHOD
- Participants
- Instruments
- Electromyography (EMG) measurement system
- Motion capture system
- Experimental protocol
- Data processing and analysis
- Movement tracking
- Muscle fatigue evaluation
- Statistical analysis
The tracking beam was located 2 m from the lateral side of the dominant arm and at the height of the elbow of each participant. An Ag-AgCl triode surface EMG sensor (Flexcomp, Thought Technology, Canada) was attached to the most prominent bulge of the biceps brachii muscle on the participant's dominant side (Hermens et al., 2000). Maximum voluntary contraction (MVC) EMG of the biceps brachii muscles was recorded to normalize EMG data during tasks.
To calculate the mean MVC EMG amplitude of the biceps brachii muscle, the middle 1s window from the EMG signal during the MVC was selected and averaged. The sagittal plane angles of the upper arm and forearm rigid bodies in the reference position were defined as 180° orientations. Two intensities of dynamic contraction tasks included: '2 kg intensity task' and '1 kg intensity task', where participants performed repeated elbow flexion and extension of the dominant arm with a given weight of dumbbells for 2 minutes.
At each intensity of the dynamic contraction tasks, the weight of the dumbbell was 1 kg and 2 kg, respectively (Figure 9). The rate of the elbow flexion and extension cycle was limited to repeat every 3 seconds (2/4, 40 bpm). The elbow joint angle was defined as the forearm flexion angle subtracted from the upper arm flexion angle.
In the isometric measurement, using the middle 4s data between each 6s measurement, the mean amplitude values of the normalized EMG (AEMG) data were calculated in the pre- and post-phase in each task. Using mean AEMG and IMNF values from each task, a second-order polynomial regression analysis was performed to develop models for predicting the progression of AEMG and IMNF magnitude with cycles. The first and last values predicted by each model in the first and last cycle (40th cycle) of each task were used to represent the phase fatigue values before and after each task.
To verify the first hypothesis and to check the significance of the development and recovery of biceps brachii muscle fatigue during each task with 2 kg and 1 kg intensity, the values before and after AEMG and central frequency (MDF and IMNF) were compared in each intensity of tasks as. dependent variables of EMG signals. To compare pre- and post-values of AEMG and central frequency from isometric measurement and dynamic measurement, one-way repeated-measures ANOVA was used at each task intensity. Also, the concordances of the trend of the change values were quantified by calculating the concordance rates of the increase and decrease of the fatigue values between the two measurement methods.
RESULTS
Movement tracking
Muscle fatigue detection
- Comparison between pre-stage and post-stage of tasks
- Effect of intensity of fatigue tasks
- Comparison of fatigue evaluation from isometric and dynamic measurements
Before and after AEMG values (top) and central frequency (bottom) at each task intensity. When evaluating fatigue from isometric and dynamic measurements, a significant effect of task intensity on changes in AEMG values and center frequency was found. AEMG change was found to increase by 27.41% MVC during 2 kg intensity tasks and decrease by 19.76% MVC during 1 kg intensity tasks.
Agreement rates for increase and decrease of AEMG between dynamic and isometric measurement methods were also varied from mean: 61.8%, standard deviation: . 11.7%). A strong positive correlation (r >.70) for the biceps brachii muscle was found between MPF change from isomeric measurements and IMNF change from dynamic measurements. Raw center frequency change values from isometric measurements (MDF) and dynamic measurements (IMNF) with the agreement rates between two measurement methods.
Raw change values of AEMG from isometric measurements and dynamic measurements with the agreement rates between two measurement methods.
DISCUSSION
Muscle fatigue detection
- Detection of development and recovery of muscle fatigue
- Validity of dynamic muscle fatigue measurements
The decrease is consistent with previous findings that the development of muscle fatigue shifts the frequency spectrum of the EMG signal to the left (Cobb & Forbes, 1923; Lippold et al., 1960). However, in this experiment it was confirmed that isometric and dynamic measurements were valid for detecting the development of biceps brachii muscle fatigue during repeated elbow flexion and extension tasks using a 2 kg dumbbell. Also, muscle fatigue could be identified midway through development even when the muscle was not fully fatigued.
The center frequency of the biceps brachii muscle was assumed to be increased after the 1 kg intensity tasks, implying recovery of muscle fatigue. The result from the current study is similar to a number of previous studies that verified the assessment of muscle fatigue recovery immediately after fatigue tasks using EMG signals. Therefore, it could be concluded that the immediate change in muscle fatigue values while performing repetitive movements is determined by the 'relative' intensity from the previous intensity, not an absolute intensity of the load.
However, in the case of evaluating muscle fatigue from AEMG, AEMG could not sensitively detect the change of fatigue value during tasks. Several studies have pointed out the difficulty in maintaining consistency during isometric muscle fatigue tests (Cifrek et al., 2009). It is also assumed that muscle fatigue evaluation from EMG amplitude was not sufficient to detect change of muscle fatigue during each set.
It is one of the main goals of this study to validate the assessment of muscle fatigue in dynamic muscle contraction situation using continuous wavelet transform compared to the traditional method of Fourier transform. Therefore, assessment of muscle fatigue was performed in both dynamic muscle contraction situation using continuous wavelet transform and isometric muscle contraction situation using Fourier transform. This implies that both methods are possible to detect the development and recovery of muscle fatigue in each 2 kg and 1 kg intensity task.
Also, the correlation result of muscle fatigue evaluation from isometric measurements and dynamic measurements shows that two fatigue detection methods have a strong positive correlation. Therefore, because dynamic measurement using continuous wavelet transform provides valid detection of the development and recovery of muscle fatigue, the method has a great advantage to detect the change of muscle fatigue without interruption of movement. Furthermore, the study was limited to verifying the evaluation of muscle fatigue in conditions of dynamic muscle contraction compared to isometric conditions.
Implication
Furthermore, considering the deviation of the change value from the center frequency, the deviation of the IMNF change between participants was smaller than the deviation of the MDF change in both 2 kg and 1 kg intensity tasks. The F-value from the ANOVA result, which compared the mean difference between the two task intensities, was much larger with IMNF change than with MDF change. It is difficult to assess the degree of sensitivity of each fatigue assessment method, but it appears that fatigue assessment in dynamic contraction situations using Continuous Wavelet Transform can quantify muscle fatigue more regularly and distinguish the effect of muscle load intensity on fatigue more powerfully than assessment of the situation of isometric muscle contraction.
Limitation
Conclusion
An investigation into the influence of a sustained contraction on the sequence of action potentials from a single motor unit. Paper presented at the 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). Discrete wavelet transform analysis of surface electromyography for the fatigue assessment of neck and shoulder muscles.
Shear wave elastography reveals different degrees of passive and active stiffness of the neck extensor muscles. Validation of the wavelet spectral estimation technique in Biceps Brachii and Brachioradialis fatigue assessment during prolonged static and low-level dynamic contractions. Evaluation of measurement strategies to increase the reliability of EMG indices for assessing back muscle fatigue and recovery.
A validation of techniques using surface EMG signals from dynamic contractions to quantify muscle fatigue during repetitive tasks. Full analysis of variance tables (compare 2kg vs 1kg intensity in change values) 2kg vs 1kg intensity in AEMG change value from isometric measurements.
