"All talk no torque"_ A novel set of metrics to quantify muscle fatigue through isometric dynamometry in Functional Electrical Stimulation (FES) muscle studies

4th International Conference on Mechanical Engineering Research

Taylor, M.J.^1,2 ; Fornusek, C.^3 ; De Chazal, P.^1,2 ; Ruys, A.J.^1

Faculty of Engineering and IT, University of Sydney, Darlington, Australia^1

Charles Perkins Centre, University of Sydney, Camperdown, Australia^2

Faculty of Health Sciences, University of Sydney, Lidcombe, Australia^3

关键词: Electrical stimulations;    Functional electrical stimulation;    Future methods;    Minimally invasive ways;    Model fatigue;    Muscle fatigues;    Peak analysis;    Qualitative assessments;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/257/1/012018/pdf
DOI  :  10.1088/1757-899X/257/1/012018

来源: IOP

【 摘 要 】

Functional Electrical Stimulation (FES) activates nerves and muscles that have been ravished and rendered paralysed by disease. As such, it is advantageous to study joint torques that arise due to electrical stimulation of muscle, to measure fatigue in an indirect, minimally-invasive way. Dynamometry is one way in which this can be achieved. In this paper, torque data is presented from an FES experiment on quadriceps, using isometric dynamometry to measure torque. A library of fatigue metrics to quantify these data are put forward. These metrics include; start and end torque peaks, percentage changes in torque over time, and maximum and minimum torque period algorithms (MTPA 1 and 2), and associated torque-time plots. It is illustrated, by example, how this novel library of metrics can model fatigue over time. Furthermore, these methods are critiqued by a qualitative assessment and compared against one another for their utility in modelling fatigue. Linear trendlines with coefficients of correlation (R2) and qualitative descriptions of data are used to achieve this. We find that although arduous, individual peak plots yield the most relevant values upon which fatigue can be assessed. Methods to calculate peaks in data have less of a utility, offset by an order of magnitude of ∼101in comparison with theoretically expected peak numbers. In light of this, we suggest that future methods would be well-inclined to investigate optimized form of peak analysis.