Fig 1.
Data from an individual participant.
A: individual traces of root-mean-square (RMS) electromyographic activity (EMG) over 50 ms prior to single pulse stimulation during contractions at each % of maximal voluntary contraction (MVC). B: superimposed and resting twitches evoked by supramaximal tibial nerve stimulation for all trials at each % of MVC. Twitches are offset vertically for clarity. Colors in panels A and B indicate corresponding trials. C: maximal M waves from soleus, medial and lateral gastrocnemius. The RMS EMG was calculated from the first phase of the maximal M wave (gray shaded regions) using the method by Thomas [20]. The lower bounds of the gray shaded regions that indicate the first phase of the M waves are not horizontal because when the signal crossed 0, the closest sampled time points are sometimes slightly above or below 0 V. D: EMG normalized to maximal EMG as a function of voluntary activation. E: EMG normalized to Mmax as a function of voluntary activation. F: EMG normalized to maximal EMG as a function of EMG normalized to Mmax. The slope of each line indicates the scaling factor for that muscle. Legend for line styles in panel D indicates corresponding muscles in panels D, E and F. Data from individual trials in these panels are shown (black circles).
Fig 2.
Non-linear relationship between EMG and torque.
Electromyographic activity (EMG) in mV and ankle torque data from all participants (n = 25, 18 males) for A: soleus (SO), B: medial gastrocnemius (MG) and C: lateral gastrocnemius (LG) muscles. Torque is normalized to the maximal torque in the maximal voluntary contractions (MVCs) performed at the start of the protocol.
Fig 3.
Effect of normalization method on the estimated level of muscle activation.
Data from each participant (n = 25, 18 males) of electromyographic activity (EMG) in each contraction level normalized to maximal EMG as a function of voluntary activation (panels A-C), EMG normalized to Mmax as a function of voluntary activation (panels D-F), and EMG normalized to maximal EMG as a function of EMG normalized to Mmax (panels G-I) for soleus (SO, panels A, D, G), medial gastrocnemius (MG, panels B, E, H) and lateral gastrocnemius (LG, panels C, F, I) muscles. Slopes of lines in panels G-I indicate scaling factors for participants and muscles.
Table 1.
Means, 95% confidence intervals (CI) and 95% prediction intervals of slopes of electromyographic activity (EMG) normalized to maximal muscle activity as a function of EMG normalized to the Mmax in the three plantarflexor muscles.
Slopes are ratios with no units.
Table 2.
Means (SD) of voluntary muscle activation (VA) and the electromyographic activity (EMG) in soleus (SO), medial gastrocnemius (MG) and lateral gastrocnemius (LG) normalized to maximal EMG (%max EMG) and normalized to Mmax (%Mmax).
EMG normalized to maximal EMG better estimates voluntary muscle activation than EMG normalized to Mmax.