Muscle fatigue assessment during robot-mediated movements

Abstract

BackgroundSeveral neuromuscular disorders present muscle fatigue as a typical symptom. Therefore, a reliable method of fatigue assessment may be crucial for understanding how specific disease features evolve over time and for developing effective rehabilitation strategies. Unfortunately, despite its importance, a standardized, reliable and objective method for fatigue measurement is lacking in clinical practice and this work investigates a practical solution.Methods40 healthy young adults performed a haptic reaching task, while holding a robotic manipulandum. Subjects were required to perform wrist flexion and extension movements in a resistive visco-elastic force field, as many times as possible, until the measured muscles (mainly flexor and extensor carpi radialis) exhibited signs of fatigue. In order to analyze the behavior and the characteristics of the two muscles, subjects were divided into two groups: in the first group, the resistive force was applied by the robot only during flexion movements, whereas, in the second group, the force was applied only during extension movements. Surface electromyographic signals (sEMG) of both flexor and extensor carpi radialis were acquired. A novel indicator to define the Onset of Fatigue (OF) was proposed and evaluated from the Mean Frequency of the sEMG signal. Furthermore, as measure of the subjects’ effort throughout the task, the energy consumption was estimated.ResultsFrom the beginning to the end of the task, as expected, all the subjects showed a decrement in Mean Frequency of the muscle involved in movements resisting the force. For the OF indicator, subjects were consistent in terms of timing of fatigue; moreover, extensor and flexor muscles presented similar OF times. The metabolic analysis showed a very low level of energy consumption and, from the behavioral point of view, the test was well tolerated by the subjects.ConclusionThe robot-aided assessment test proposed in this study, proved to be an easy to administer, fast and reliable method for objectively measuring muscular fatigue in a healthy population. This work developed a framework for an evaluation that can be deployed in a clinical practice with patients presenting neuromuscular disorders. Considering the low metabolic demand, the requested effort would likely be well tolerated by clinical populations.

Coupling Robot-Aided Assessment and Surface Electromyography (sEMG) to Evaluate the Effect of Muscle Fatigue on Wrist Position Sense in the Flexion-Extension Plane

“…In recent years, collaborative work by our laboratories has focused on developing and testing robotic devices for upper extremity neurorehabilitation in healthy (Mugnosso et al, 2018), and neurologically impaired populations (Squeri et al, 2011; De Santis et al, 2015; Marini et al, 2017b). The advantage of robotic devices is that they have better diagnostic and prognostic precision than current clinical evaluation measures, resulting in a greater sensitivity to subtle differences in neurological status (Rinderknecht et al, 2016).…”

Dynamic Wrist Flexion and Extension Fatigue Induced via Submaximal Contractions Similarly Impairs Hand Tracking Accuracy in Young Adult Males and Females

“…Finally, the addition of electromyographic measurements would have been useful in examining the extent of muscular fatigue through spectral analysis, and in activation pattern changes. Additionally, in the future, we will be able to employ spectral analysis, the Dimitrov Index and root mean square amplitude of the electromyographic signal in combination with the WristBot to compute the onset of fatigue (Mugnosso et al, 2018 ).…”

sEMG-Based Detection of Compensation Caused by Fatigue During Rehabilitation Therapy: A Pilot Study

“…The raw sEMG signals were first bandpass filtered with cut-off frequencies of 10 Hz and 500 Hz and notch filtered at 50 Hz to reduce the noise content, as recommended by the European SENIAM project [36]. We used an overlap analysis window with a window length of 512 ms and a window sliding step size of 256 ms. Because the training was repetitive, the MDF during the active state could reflect the real muscle fatigue state [37]. Thus, the sEMG signal data were segmented to focus the analysis on the active segments.…”