Torque (Figure 1B), i.e. when the physique leaned forward from its equilibrium position the plantarflexion torque increased (a lot more adverse). Conversely, muscle activations (EMG envelopes in Figure 1C-E) were modulated roughly in phase with postural sway. In the simulations, TA muscle was silent through postural sway (not shown). A quantitative analysis was performed to validate the model with respect for the readily available data from the literature. Common timedomain metrics were calculated in the COP time series and compared to information from standard subjects and vestibular loss individuals standing on a force plate without the need of visual information (see Table 1). Root mean square (RMS) and mean velocity (MV) of simulated COP were greater than the values observed experimentally in normal subjects, but compatible with data from vestibular loss individuals. Yet another quantitative validation was depending on a crosscorrelation ML364 web evaluation performed in between the COM and COP time series (Figure 2A-B), also as among COP and EMG envelopes (Figure 2C-D). COM and COP had been highly correlated (r 1) at lag zero. COP and EMG envelopes have been positively correlated with all the correlation peak occurring at a optimistic lag. Correlation coefficients (r) and cross-correlation peak lag values were compatible with PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/20176980 experimental data from healthful subjects (see Table 1). Normally, correlation coefficients have been higher for Gastrocnemii in comparison to the SO, and muscles’ activations (EMGs) were advanced by roughly 20000 ms in relationPLOS Computational Biology | www.ploscompbiol.orgto the postural sway (COP). The 50 energy frequency (F 50) estimated from the COP power spectrum (see Figure 2E-F) resulted very similar to the value from healthy subjects (see Table 1). COP power spectra of both model structures had been limited to 1 Hz. A final quantitative validation was depending on the pooled histogram of COM displacements (1-mm bins) as shown in Figure three (data are from the simulations of Model 2). The histogram shape was bimodal, with two peaks around the equilibrium position of the inverted pendulum (value 0 within the abscissa). The Jarque-Bera goodness-of-fit test was applied to verify if this histogram might be fitted by a typical Gaussian probability density function [11]. The null-hypothesis (the histogram comes from an unimodal Gaussian function) was rejected (p 0:001). The same outcome was obtained for Model 1.Intermittent Recruitment with the Motor UnitsFigures four and 5 show how the spike trains from spinal MNs, INs, and afferent fibres have been modulated in the course of postural sway. An exciting qualitative discovering was that MUs from the MG muscle were intermittently recruited/de-recruited as the inverted pendulum swayed forward/backward (Figure 4B). This intermittent pattern of MU recruitment was similar for the LG muscle (not shown), but much less evident for the SO muscle (see Figure 5A).Cross-correlation functions and centre of pressure (COP) energy spectra for common simulations carried out on Model 1 (graphs A, C, and E) and Model two (graphs B, D, and F). (A-B) Cross-correlation functions involving centre of mass (COM) and COP. Note that for each models, cross-correlation peaks occurred at zero lag (dashed lines). (C-D) Cross-correlation functions involving COP and muscle electromyograms (EMGs). Black, red, and blue curves represent cross-correlation functions for Soleus (SO), Medial Gastrocnemius (MG), and Lateral Gastrocnemius (LG), respectively. Irrespective from the model structure, there was a lag o.
Posted inUncategorized