Any movement performed by the human body implies a certain muscular effort, however small it may be. Whether it is a contraction, concentric, eccentric, or isometric, the muscle generates electrical potentials that are studied through electromyography (EMG) (Konrad, 2005). Measuring on the surface of the skin, known as surface electromyography (sEMG), has particular interest for rehabilitation, sports, and even ergonomics (Vigotsky, Halperin, Lehman, Trajano, & Vieira, 2018).

The acquisition of the EMG signal can be made through conductor electrodes or conductive fabric that must be applied to the groups of muscles to be measured. Within the context of football, it is necessary to take into account the comfort of the athlete on choosing the most appropriate technology and, above all, to focus on the muscles that might be more participative in the task: the lower limbs. A few solutions have been made available in the market for this purpose, as is the case of Myontec MBody 3, which allows collecting multiple channels of muscle activity in the legs. The muscle activity of a given player i can be quantified by the amplitude value of the signal, A?' [I], of a given channel ch out of a total number of channels chtol.

Yet, even if multiple channels might be available, most of the time, additional features still need to be computed out of EMG data, employing pre-processing routines, including filtering, rectification, and smoothing (Merletti & Di Torino, 1999). Filtering is intended to remove unwanted noise from the original signal. In some cases, when the acquisition device returns clean signals, this process is not necessary. However, when there is a need to filter, a band-pass filter is usually applied, removing both low and high frequencies. The removal of the low frequencies eliminates the baseline deviation which is usually associated with slight movements and even breathing, with typical frequency values of 5-20 Hz. The high-frequency cut-off prevents the occurrence of signal aliasing, with typical values between 200 Hz and 1 kHz. With the rectification, a reorganization of the signal is made to calculate standard amplitude parameters, such as average, peak values, and area. Basically, all negative amplitudes are converted to positive amplitudes. It should be noted that, unlike filtering, this procedure does not affect signal noise, so smoothing may be necessary. Through smoothing, it is possible to create a linear envelope of the signal, leaving only a centred part of it. Usually, a Butterworth filter is used, which is a low-pass filter, being considered one of the best digital filters to decrease the relationship between signal and noise (Mello, Oliveira, & Nadal, 2007).

< Prev   CONTENTS   Source   Next >