Evaluation of lower limb muscle fatigue in 400-meters sprinting using three pacing strategies based on surface EMG signals

Muscle fatigue in sports science is an established research area where various techniques and types of muscles have been studied in order to understand the fatigue condition during running 400-meters. Muscle fatigue can be used as an indicator for predicting muscle injury and other muscle problems which can decrease athletes’ performance during running. Where, the improper running strategy may lead to discomfort and injuries, especially for non-athletes. For this reason, it is necessary to develop the strategy of running to prevent the muscles fatigue in early time during running 400-meters. The main aim of this research work is to evaluate the muscle fatigue of Rectus Femoris (RF), Biceps Femoris (BF), Gluteus Maximus (GM), Gastrocnemius Lateralis (GL), and Gastrocnemius Medialis (GMS) muscles of the right leg during running 400-meters with three types of running strategies. The 1st strategy was the 1st 200 meters, running at 87% - 94% of the full speed and the last 200 meters sprinting (full speed). The 2nd strategy was the 1st 300 meters, running at 87% - 94% of sprinting and the last 100 meters sprinting. The 3rd strategy was running at 87% - 94% of sprinting for 400-meters. The Surface Electromyography (sEMG) signals have been collected from 15 active male healthy and normal subjects (non-athletes) by using bipolar electrodes from the right lower extremity’s muscles during the run on the tartan athletics track, then the sEMG signals were processed and one-way ANOVA was applied for the raw data of each muscle separately to find out the significance for each muscle between the running strategies. Next, the processed EMG signals were transformed to the frequency domain using Fast Fourier Transform (FFT) and to the time-frequency domain using Short Time Fourier Transform (STFT). The extracted features methods used in this research were the root mean square (RMS), mean absolute value (MAV), mean frequency (MNF), median frequency (MDF), instantaneous median frequency (IMDF), and instantaneous mean frequency (IMNF). Next, the linear and polynomial (quadratic 2nd and cubic 3rd order) regression were applied to these features separately to detect the muscle fatigue at each 25-meters of 400-meters with three running strategies to know which one gets less fatigue and gets higher fatigue, based on the accuracy of the regression model. To enhance the performance of the regression model for detecting the muscle fatigue, new approaches have been proposed, namely mean of RMS and MAV (MRM), mean of MDF and MNF (MDNF), mean of IMDF and IMNF (MIDINF), standard deviation of RMS and MAV (SDRM), standard deviation of MDF and MNF (SDDNF), and standard deviation of IMDF and IMNF (SDIDINF). From the results, it has been observed that the higher accuracy was with a cubic regression model of the MRM (R-value was 0.836) and MDNF (R-value was 0.933) features, where the lowest fatigue index was during running with 1st strategy and the highest fatigue index was during running with the 2nd strategy for most the selected muscles. GM and GL muscles got less fatigue based on Joint Analysis of EMG Spectrum and Amplitude (JASA) method. Where the fatigue index of GM muscle was 0.054 and -0.826, and for GL muscle it was 0.026 and -0.774 with MRM and MDNF features respectively. It can be concluded that, the cubic regression model is the better to fit the data and the 1st strategy is better for the runners to prevent them from getting earlier fatigues during running.