|
| |
 |
|
| ORIGINAL ARTICLE |
|
| Year : 2019 | Volume
: 62
| Issue : 4 | Page : 166-174 |
|
The effect of visuospatial resolution on discharge variability among motor units and force–discharge relation
Yi-Ching Chen1, Chia-Li Shih2, Yen-Ting Lin3, Ing-Shiou Hwang4
1 Department of Physical Therapy, Chung Shan Medical University; Physical Therapy Room, Chung Shan Medical University Hospital, Taichung City, Taiwan
2 Department of Rehabilitation Medicine, Tainan Municipal An-Nan Hospital, Tainan, Taiwan
3 Physical Education Office, Asian University, Taichung City, Taiwan
4 Institute of Allied Health Sciences; Department of Physical Therapy, College of Medicine, National Cheng Kung University, Tainan City, Taiwan
| Date of Submission | 31-Jan-2019 |
| Date of Decision | 15-Jun-2019 |
| Date of Acceptance | 19-Jun-2019 |
| Date of Web Publication | 19-Aug-2019 |
Correspondence Address:
Prof. Ing-Shiou Hwang
Institute of Allied Health Sciences, College of Medicine, National Cheng Kung University, Tainan City
Taiwan
 Source of Support: None, Conflict of Interest: None
DOI: 10.4103/CJP.CJP_12_19
Although force steadiness varies with visuospatial information, accountable motor unit (MU) behaviors are not fully understood. This study investigated the modulation of MU discharges and force–discharge relation due to variations in the spatial resolution of visual feedback, with a particular focus on discharge variability among MUs. Fourteen young adults produced isometric force at 10% of maximal voluntary contraction (MVC) through index abduction, under the conditions of force trajectory displayed with low visual gain (LVG) and high visual gain (HVG). Together with smaller and more complex force fluctuations, HVG resulted in greater variabilities of the mean interspike interval and discharge irregularity among MUs than LVG did. Estimated via smoothening of a cumulative spike train of all MUs, global discharge rate was tuned to visual gain, with a more complex global discharge rate and a lower force–discharge relation in the HVG condition. These higher discharge variabilities were linked to larger variance of the common drive received by MUs for regulation of muscle force with higher visuospatial information. In summary, higher visuospatial information improves force steadiness with more complex force fluctuations, underlying joint effects of low-pass filter property of the musculotendon complex and central modulation of discharge variability among MUs.
Keywords: Electromyography, isometric contraction, motor unit, visuomotor
How to cite this article:
Chen YC, Shih CL, Lin YT, Hwang IS. The effect of visuospatial resolution on discharge variability among motor units and force–discharge relation. Chin J Physiol 2019;62:166-74 |
How to cite this URL:
Chen YC, Shih CL, Lin YT, Hwang IS. The effect of visuospatial resolution on discharge variability among motor units and force–discharge relation. Chin J Physiol [serial online] 2019 [cited 2023 Nov 30];62:166-74. Available from: https://www.cjphysiology.org/text.asp?2019/62/4/166/264755 |
| Introduction | |  |
The spatial properties of visual information are keyed to daily force tasks under visual guidance.[1] If the spatial resolution of visual feedback on force output is high (or a smaller change in force occupies a greater number of pixels on the screen), mismatches between the target and force signals are multiplied.[2] A more easily distinguishable image of execution errors on the retina could alter neural firing in the visual cortex and visual object perception,[3] prior to visuomotor corrections. The effect of visual gain on force steadiness (or force fluctuations) has been shown to be more prominent at lower exertion levels (<4 N for isometric index abduction).[1],[4] Except for excessively high visual gain (HVG) (or pixel/force ratio),[1] displaying force output with HVG in the effective range always leads to superior task accuracy, in conjunction with smaller but more complex force fluctuations, than displaying it with low visual gain (LVG).[5],[6],[7] Under the framework of the sampled feedback model, task improvement due to the use of HVG could be attributable to better use of spatial information within the visual feedback to enrich the force gradation strategy,[8],[9],[10] as low-frequency force fluctuations are considered to be a number of corrective attempts to achieve target exactness, known as an additive control mechanism of task accuracy.[11],[12] In effect, force regulation with visual gain feedback must involve organization of the neural drive to motor units (MUs) from the motor cortex via a distributed visuomotor system.[13],[14] However, neurophysiological mechanisms responsible for dimensional changes in force fluctuations with visual gain are not completely clear.
Relatively few studies have been undertaken to investigate the visual gain effect on MU behaviors. For young adults, previous studies have shown that the mean and variability of the interspike interval (ISIi) of a single MU are not modified by visual gain.[10],[15] Although MU discharge times became less regular for the highest visual gains as compared to those with no visual feedback, the regularity of a single MU discharge in the low-gain and high-gain conditions appears not to differ in the temporal domain.[10],[15] Schmied et al. reported only a slight increase in the synchronous probability of MU discharges under the high-gain condition.[16] The significant impact of visual gain on MU behaviors was first reported by Laine et al. who noted that high-gain feedback led to lower coherence (1–5 Hz) but larger coherence at high-frequency coherence spectra (6–25 Hz) among pairs of MUs.[17] Despite these reported facts, no past MU studies could adequately account for the variations in force fluctuations with respect to visual gain. The relation of force fluctuations and MU discharges under the various visual gain conditions was far from clear. One major constraint of the previous studies was that comparatively fewer MUs were investigated, whereas the visual gain effect on force fluctuations was likely to be reconciled by discharge variability among a population of MUs, rather than by averaged discharge properties from a few MUs. In addition, previous visual gain work focused on discharge variability within an MU such as coefficient of MU spike train,[15] rather than discharge variability properties among MUs.
Using multielectrode surface electromyography (EMG) and mathematical decomposition techniques,[18],[19] this study aimed to investigate the effect of visual gain on MU discharges, with a particular focus on discharge variability among MUs, which could be susceptible to visuomotor integration and force steadiness maintenance at the periphery. In addition, this study contrasted variations in the common drive to MUs and the global discharge rate from cumulative spike trains of MUs between the LVG and HVG feedbacks. Our main hypotheses were that (1) HVG feedback would add to discharge variations among MUs, so as to alter the global discharge rate and its representation to low-frequency force fluctuations and (2) variation in the common drive would increase with the use of HVG feedback.
| Materials and Methods | | |
The participants were 14 healthy adults (6 males and 8 females; mean age: 25.8 ± 6.8 years). All were self-reported as being right handed, and none had symptoms or signs of neuromuscular diseases. The research project was approved by an authorized institutional human research review board (institutional review board) at the University Hospital of the National Cheng Kung University, Taiwan. All experiments were performed in accordance with relevant guidelines and regulations. All the participants signed informed consents before the experiment, conforming to the Declaration of Helsinki.
Experimental procedures
All participants completed a unilateral static force task (isometric index abduction) with the dominant hand under two visual conditions (LVG and HVG) at a low force level (10% MVC). In addition to minimizing the overlapping of MU action potentials (MUAPs), the low force level in this study was used to highlight the visual gain effect on force characteristics and MU discharges.[5] The participant was seated with his/her palm and forearm firmly fixed within a thermoplastic splint on a table. The index finger was held slightly abducted (5°), and its abduction force was measured using a force transducer (Model: MB-100, Interface Inc. Scottsdale, AZ, USA) [Figure 1]. Prior to the experiment, the MVC of the first dorsal interosseus (FDI) was predetermined as the peak force during three MVCs of 3-s duration separated by 3-min pauses. The participants were instructed to couple the force output of the index abduction to a target force line (10% MVC) displayed on a 22-inch monitor (1920 × 1080 pixels). After a 3-s relaxation period, the participant surged the force output from zero to the target level within 1 s (the up ramp to the force target: 10% MVC/second) and maintained steady contraction for another 12 s under visual guidance [Figure 1]. The force output decreased to zero (the down ramp to the rest: −10% MVC/second) for the rest of the experimental trial. Each experimental trial took 18 s to complete. The force feedback was displayed on the monitor in two different modes of spatial resolution. In the LVG condition, the visual gain for the display force output was set at 4.8 pixels per 1% MVC, and in the HVG condition, 48 pixels per 1% MVC. The HVG was not so high that it would destabilize force output because force variability exhibits a U-shaped trend as visual gain progressively increases.[1] The participants completed three practice trials to learn how to control the force in the HVG and LVG conditions before the experiment. Then, they performed the experimental trial three times for each visual condition, and the order of the LVG and HVG trials was randomized. An intertrial interval of rest of 3 min was used to prevent a potential fatigue effect. The force signal was sampled at 1 kHz by an analog-to-digital converter with 16-bit resolution (DAQCard-6024E; National Instruments Inc. Austin, TX, USA), controlled by a custom program on a LabVIEW platform (LabVIEW v. 8.5, National Instruments Inc. Austin, TX, USA). | Figure 1: Experimental apparatus (left) and physiological data in the time window of interest (the 5th–15th s) (right)
Click here to view |
Electrophysiological recordings
Synchronized with the force signal, multielectrode surface EMG (Bagnoli sEMG system, Delsys Inc., Natick, MA, USA) with five surface pin-sensors (0.5-mm diameter) was used to record the activities of the FDI muscle. The pin-sensors were placed at the center and at the corners of a 5 mm × 5 mm square. Protruding from the housing of the electrode, the pin-sensors with blunted ends did not puncture the skin when they were pressed against the skin to detect muscle activities. The analog EMG signals from each pin-sensor were amplified (gain = 1000) and band-pass filtered (cutoff frequencies 20 and 450 Hz, 80 dB/octave roll-off).[16] The sampling rate of each channel for decomposition-based surface EMG was 20 kHz.[19],[20] Pair-wise subtractions of voltages (four single differential EMG channels) at the five pin-detections derived multichannel surface EMG signals. This spatial filtering could improve the capacity of the subsequent decomposition procedure of the single MUAP. To ensure decomposition quality, the signal-to-noise ratio of the EMG signal was strictly controlled with baseline noise under a peak-to-peak value of 20 μV.[19],[20] The software used for the postdecomposition processing of action potential morphology was EMGWorks v. 4.1 (Delsys Inc. Natick, MA, USA), which extracted action potential “templates” of as many MUAP trains (MUAPTs) as practically possible according to a previous proof of principle.[19],[20] Signal regions with the extracted templates, superposed with each other or previously unidentified, were searched and decomposed.[19],[21] The computation algorithm had repeatedly been shown to produce convincing decomposition results that discriminate overlapping action potentials for static isometric contraction [22],[23] via independent verification methods.[18] All the MUs identified with the decomposition software were included for discharge pattern analysis. The accuracy of the EMG decomposition algorithm of the Bagnoli sEMG system was evaluated with the Decomposition-Synthesis-Decomposition-Compare (DSDC) test.[20],[22] The degree of accuracy for each MUAPT was assessed with the total number of unmatched events and the total number of true events by comparing the firing instances of the actual composed EMG signal and the decomposed synthesized signals. The overall accuracy of the decomposition was obtained by averaging the individual accuracies of each MUAPT in an experimental trial across the participants. The average decomposition accuracy across all the participants was 93.6% ± 1.9% in this study, in good agreement with previous studies that used the same approach.[20],[22]
Data analysis
The force signal of index abduction was first conditioned with a low-pass filter (cutoff frequency: 6 Hz) to exclude involuntary tremulous movements unrelated to corrective actions.[24],[25] Previous studies have also reported that most of the effects of visuomotor processes are in the 0–4 Hz band of force output [9],[10] and that high-frequency force components, such as 8–12 Hz physiological tremor, are functionally independent of force-tuning strategies.[11],[12] Considering the stability of force generation, only force data from the 5th to 15th s were further analyzed [Figure 1]. Three variables from the truncated force data were calculated, including standard deviations, sample entropy (SampEn), and the mean frequency (MF) of force output after removal of a linear trend of force fluctuations. In accordance with most previous studies,[1],[5] the force variables in this study were performed on down-sampled force fluctuations at 100 Hz. The down-sampling process was helpful to reduce the computational load, and the time scale of 10 ms for entropy measures of isometric force data has been commonly used in previous studies.[1],[5],[8] Granting that force fluctuations during static isometric contraction were stochastic in nature,[5] the complexity of force fluctuations was quantified with SampEn, which measures the logarithmic likelihood that runs of patterns that are close to each other for m observations (within tolerance r) will remain close on subsequent incremental comparisons. The mathematical formula of SampEn is as follows:
Where r = 20% of the standard deviation of the force channel, m is the length of the template (m = 2), and N is the number of data points in the time series. Ai is the number of matches of the ith template of length m + 1 data points, and Bi is the number of matches of the ith template of length m data points. A lower value represents greater regularity of force characteristics. A larger SampEn represents a more complex temporal structure of the low-frequency force fluctuations. The MF of low-frequency force fluctuations was determined through the force spectral profile, which was estimated using a fast Fourier transform and the Welch method (Hanning window, window length: 3 s, overlapping time segment: ¼ × window length) with a spectral resolution of 0.1 Hz. The MF of the reflected spectral shift of force fluctuations and a higher MF functionally implied more corrective attempts during the force gradation process.[8],[12],[26]
With decomposition processing, the binary spike trains that coded activation for all MUs with a value of 0 or 1 in an 18-s experimental trial were identified [Figure 2]. In line with the truncated force data, the discharge variables were determined in the time window of interest (the 5th–15th s) based on the decomposed EMG data of 18 s. The validity of the EMG decomposition of each MUAPT was evaluated with the DSDC test.[20],[27] MUs of low decomposition accuracy (<90%) were excluded from the analysis. Time intervals of a MU spike train were averaged to obtain the mean ISIi of a single MU (M-ISI). M-ISI of all the identified MUs in an experimental trial was averaged to obtain the grand average of mean ISI (M-ISIGAV). The experimentally observed ISI irregularity of a single MU was assessed with the irregularity index (IR), which was resistant to variations in the length of data analyzed across trials.[28],[29] Given a series of ISIi in a MU spike train, IR was formulated as  . The irregularity of MU discharges was defined as the population mean of IR for all MUs in an experimental trial, or the grand average of ISI IR (ISI_IRGAV). Finally, coefficient of variation (CV) of ISI_IR was determined to highlight variations in discharge irregularity among MUs in an experimental trial (CV of ISI_IR) [Figure 2]. For an individual MU, time series of discharge rate was obtained by a smoothening procedure using a 400-ms Hanning window to suppress tremulous discharges above 8 Hz.[27],[30] To provide an estimate of the common modulation of the discharge rate, we performed common drive analysis on 10-s epochs of smoothed discharge data from all MU pairs.[31],[32] The traditional common drive index (CDI) and variation in discharge commonality (CDI_CV) in an experimental trial were defined as the grand average and CV of the maximum values of normalized cross-correlation plots of the smoothed discharge rate within 50 ms from all MU pairs. The cross-correlation plot  was mathematically the smoothed cross-correlation function.  , where  , x and y are smoothed discharge rates of the two considered MUs, t is the time variable, and τ is the time-lag variable. For variance stabilization, the traditional CDI and CDI_CV were remapped to normalized CDI (N-CDI) and normalized CDI_CV (N-CDI_CV) with Fisher z transformation, respectively. . Given that force exerted by a muscle is a function of the discharge pattern of all activated MUs, we spatially summated the MU spike trains from all the MUs to obtain a cumulative spike train in an experimental trial. The global discharge rate was defined by smoothening the cumulative spike trains by convoluting a 400-ms Hanning window because force steadiness is exclusively determined by low-pass filtered components of the neural drive to the muscle [33],[34] [Figure 2]. Following linear trend removal and a down-sampling process to 100 Hz, the size and complexity of the variability for the conditioned global discharge rate were characterized by CV and multiscale entropy (MSE) [Figure 2]. The merit of MSE is that it can quantify the degree of complexity within a time series across multiple time scales based on SampEn (m = 2 and r = 20% of the standard deviation of the conditioned global discharge rate) and a coarse-graining procedure.[35] Hence, we could better characterize the complexity of the global discharge rate with the MSE area, which equaled the sum of SampEn values across the time scales from 1 to 10 (each time scale = 10 ms). The relationships of force and discharge relation were assessed linearly with the peak value of cross-correlation (Xcorr) of the force fluctuations and global discharge after the removal of a linear trend. Because the viscous resistances of the musculotendon system may attenuate the high-frequency force components, the nonlinearity of the force–discharge relation was also assessed with entropy-based mutual information (MI).[36] MI (MI [X; Y]) is mathematically defined as:  where p (x, y) is the joint probability density function of force fluctuations (X) and detrended mean smoothed discharge rate trace (Y), and p1(x) and p2(y) are the marginal probability density functions of the two time series. | Figure 2: Schematic illustration for acquisition of discharge properties, discharge variability among motor units, common drive index, and global discharge rate
Click here to view |
All force variables and discharge variables of the three experimental trials were calculated participant by participant and then averaged across participants in the LVG and HVG conditions. The major interest of this study was to examine the visual gain effect (LVG vs. HVG) on the low-frequency force fluctuations, MU discharge patterns, and coupling between the force fluctuations and global discharge rate. The average number of MUs in an experimental trial between the two visual conditions was examined with paired t statistics. Hotelling's T-squared statistics were used to contrast the force variables (standard deviations, SampEn, and MF), grand average of discharge variables (M-ISIGAV and ISI_IRGAV), CV in discharge variables among MUs (CV of M-ISI and CV of ISI_IR), global discharge characteristics (CV and MSE area), CDI variables (N-CDI and N-CDI_CV), and force–discharge relations (peak Xcorr and MI between force fluctuations and global discharge rate). The level of significance for Hotelling's T-squared statistics was 0.05. The post hoc test was the Simes test,[37] which would not produce overcorrection like the Bonferroni test. For all post hoc hypotheses  , the Simes test did not reject elementary Hi if pi≤ i × 0.05/m for ordered unadjusted P values (p1≤ … ≤ pm). Type 1 error rate using the Simes test was proved to be exactly 0.05 when elementary hypotheses were independent. Signal processing and statistical analyses were completed using Matlab (Mathworks Inc., Natick, MA, USA) and the Statistical Package for Social Sciences (SPSS) for Windows v. 17.0 (SPSS Inc., Armonk, NY, USA). Data reported in the text, figure, and tables without specific illustrations are presented as mean ± standard deviation.
| Results | | |
The mean value of MVC of all participants was 23.59 ± 8.11 N (16.15–40.06 N). The target force of 10% MVC was proved to be sensitive to visual gain effect. [Table 1] summarizes the results of Hotelling's T-squared statistics on differences in force fluctuation variables (MF, SampEn, and standard deviation) between the LVG and HVG conditions. Force fluctuations were subject to visual gain manipulation (P < 0.001). Post hoc analysis using the Simes test revealed that the HVG condition exhibited higher MF (P < 0.0167), higher SampEn (P < 0.033), and lower standard deviation (P < 0.05) than the LVG condition did. | Table 1: The contrast of force fluctuation properties between low-force isometric contraction of index abduction with low visual gains and high visual gains
Click here to view |
The average number of decomposed MUs in an experimental trial did not differ for the LVG and HVG conditions (LVG: 19.69 ± 2.15, range: 15.3–22.0; HVG: 20.12 ± 2.97, range: 13.7–23.0; t13= −0.691, P = 0.502). [Table 2] summarizes the grand averages of discharge variables and CV for discharge variables among MUs in the LVG and HVG conditions. The results of Hotelling's T-squared statistics revealed that the grand averages of the discharge variables were independent of visual gain manipulation (P > 0.05). Hence, averaged discharge rate and discharge irregularity within an MU were not sensitive to visual gain. In contrast, CV for discharge variables among MUs was a function of visual gain (P = 0.038). The Simes test further revealed that the CV for mean ISIi(CV of M-ISI) across MUs (P = 0.013) and CV of the mean IR of ISIi(CV of ISI_IR) across MUs (P = 0.049) were larger in the HVG condition than in the LVG condition. These observations implied that HVG enhanced discharge variability among MUs during static force tracking. | Table 2: The contrast motor unit discharge variables in the low visual gain and high visual gain conditions
Click here to view |
In terms of Z values, [Figure 3]a shows normalized cross-correlation plots of the smoothed discharge rates for all MU pairs in a typical trial of the LVG and HVG conditions. The normalized CDI (N-CDI) was the mean value of all the cross-correlation peaks around zero lag time. The variation in normalized discharge commonality (N-CDI_CV) was the CV of the normalized cross-correlation peaks for all MU pairs, which represented strength scattering of the N-CDI in an experimental trial. [Figure 3]b contrasts the population means of the N-CDI and N-CDI_CV between the two visual conditions. The Hotelling's T-squared statistics revealed a significant difference in CDI variables of the two conditions (Λ3,13=0.522, P = 0.02). The N-CDI_CV was larger in the HVG condition than in the LVG condition (P =0.004), whereas the N-CDI was not different for the two visual conditions (P = 0.120). [Table 3] shows the population mean and standard deviations of the CV and MSE area of the global discharge rate in the LVG and HVG conditions. Overall, the properties of the global discharge rate varied with visual gain (P < 0.001), and the Simes test revealed that the HVG condition exhibited a larger CV (P < 0.025) and larger MSE area (P < 0.05) than the LVG condition. | Figure 3: Normalized common drive index and normalized variation in discharge commonality for the conditions of low and high visual gain. (a) Typical cross-correlation plots. (b) The contrast population means of normalized common drive index and normalized variation in discharge commonality between two visual conditions
Click here to view |
 | Table 3: The contrast of the variability structure of the global discharge rate in the low visual gain and high visual gain conditions
Click here to view |
The relationships between force fluctuations and the detrended global discharge rates (or force–discharge relation) were characterized with peak cross-correlation (Xcorr) and MI, respectively [Figure 4]. Population means of Xcorr were 0.364 ± 0.058 and 0.333 ± 0.023 for the HVG and LVG conditions, respectively. Population means of MI were 0.184 ± 0.023 and 0.161 ± 0.030 for the HVG and LVG conditions, respectively. The force–discharge relations were significantly dependent on visual gain (Λ3,13= 0.417, P = 0.005). The Simes test revealed that nonlinear force–discharge relations in terms of MI (P = 0.026) were marginally smaller in the HVG condition than in the LVG condition. The linear force–discharge relations in terms of Xcorr were also smaller in the HVG condition than in the LVG condition (P = 0.043). | Figure 4: Visual gain effect on peak cross correlation (Xcorr) and mutual information between force fluctuations and global motor unit discharge
Click here to view |
| Discussion | | |
The novel finding of this study was that HVG appropriately enhanced the discharge variability among MUs, variability in discharge commonality, and the complexity of the global discharge rate. Next, HVG undermined the association between the low-frequency force fluctuation and global discharge rate.
At the behavior level, there was a close similarity between our present results and results of previous studies,[2],[5],[7],[10] considering the structural changes in low-frequency force fluctuations with visual gain at a low force level. According to the sampled process model,[11] low-frequency fluctuations come from the intermittent use of the visual feedback information so that the visuomotor system can more effectively correct force-tracking deviations without being severely constrained by inherent feedback delays. Several previous studies have emphasized the role of low-frequency force fluctuations in the accuracy control mechanism.[8],[12],[25] In view of the smaller and more complex force fluctuations with higher MF [Table 1], the participants tended to use a richer corrective strategy to maintain force steadiness via frequent and tiny adjustments to achieve target exactness in the HVG condition.[6],[38] Although force behaviors were gain dependent [Table 1], we argue that visual-gain dependent force fluctuation properties could not be ascribed to variations in the degree of MU synchronization [16] because MU synchronization often adds to force regularity and the size of force fluctuations,[39],[40],[41] opposing the smaller force fluctuations with more complexity in the HVG condition.
Intriguingly, no known MU behaviors could well explain the observed visual adaptations in force microstructures, although several lines of neural evidence have shown reciprocal connections among the sensorimotor network, including the visual cortex (V3 and V5), superior parietal lobule, and precentral motor areas.[7],[13],[42] MU discharge should have selectively corresponded to visuospatial information of a force task. Akin to previous observations,[10],[15],[16] the discharge characteristics of a single MU did not significantly vary with visual gain [Table 2]. However, it was particularly worthy of noting significant increases in the CVs of the mean discharge rate and discharge irregularity among MUs under the HVG condition with ample visuospatial information [Table 2]. The facts indicated that HVG added to discharge variability in mean ISIi and discharge of ISIi among MUs for an isometric force task at a low exertion level, a physiological phenomenon that has yet to be inferred through detailed analysis. The increased discharge variability among MUs contributed to increases in the degree of complexity of the global discharge rate [Table 3], providing a sound explanation of the force fluctuations with higher MF and greater complexity in the HVG condition [Table 1]. The potentiation of discharge variability among MUs resulted from the MUs receiving a common synaptic input with larger variances (N-CDI_CV) [Figure 3]b, when spatial information within visual feedback multiplied. Discharge commonality was likely tuned to Ia afferent feedback at the spinal level, as Laine et al. reported a larger H-reflex amplitude concurrent with spectral changes in common input to motor neurons with the use of high-sensitivity spatial feedback.[17] As discharge probability depends on both Ia presynaptic and postsynaptic mechanisms, motoneurons of the human hand could be modulated by cortico-motoneuronal and peripheral-segmental Ia excitatory inputs due to visual attention changes.[43] Functionally, the increases in complexity in the global discharge rate and discharge variability among MUs imply an increase in neural noise. The enhanced neural noise resembles the addition of discharge randomness to the motoneuronal pool.[25] In effect, addition of optimal amounts of noise to a nonlinear neural network provides processing benefits for target discrimination,[44],[45] according to stochastic resonance theory. Central to this interpretation, HVG that exhibited enhanced discharge variability among MUs in the motoneuronal pool could benefit from stochastic resonance for superior force gradation with richer complexity. Despite additional neural noise, the summated twitch force above 1 Hz in the HVG condition could be condensed by the low-pass filter property of the musculotendon complex.[46] As a consequence, it was not surprising to observe reductions in linear and nonlinear force–discharge relations in the HVG condition [Figure 4], in that a portion of high-frequency neural drive to muscle was dampened. Nevertheless, we could not preclude that the waning of force–discharge relation in the HVG condition was alternatively a consequence of an increase in activity from synergistic or antagonistic muscles that contributed to the recorded force.
This study used decomposition-based surface EMG to identify the discharge events of individual MUs. The mathematical decomposition approach, based on either blind source separation [47],[48] or artificial intelligence,[20] has been shown to successfully extract the spike trains of individual MU discharges from EMG interference patterns without being constrained by muscle architectures and nonphysiological parameters. For the human FDI muscle at low levels of muscle contraction, the present surface EMG decomposition algorithm produced spike timings similar to those of the intramuscular recording.[21] Although the accuracy of different surface EMG decomposition approaches is still debatable,[22],[49] it is not possible to resolve the preexisting disputes over surface EMG decomposition in this study. To be prudent, we considered the MU decomposition at a low exertion level (around 10% MVC) with less superposition of MUAPs. We also applied the “reconstruct-and-test” procedure to validate the accuracy of the obtained identifications.[22] In practical reality, the high decomposition accuracy in this study (93.6% ± 1.9%) was well consistent with previous reports using the same decomposition algorithm (91.50%–96.65%).[19],[20] Unlike intramuscular recording, which is confined to a smaller territory of the pickup area, decomposition-based surface EMG detects more active MUs during isometric contraction,[50],[51] so we were better able to assess the discharge variability among MUs or the global discharge rate. The population variability of MU discharges was not characterized in past studies based on fewer identifiable MUs.[10],[15],[17] Finally, it was a technical compromise to minimize possible visual adaptation with a relatively shorter period (12 s) for a static isometric contraction. Together with EMG signal decomposition technology that allows for the observation of a greater number of active MUs, the discharge behaviors of the FDI muscle reported in this study were actually larger than those of previous studies using invasive approach.[27],[52],[53] The degree of the mean CDI of the FDI muscle was still compatible with the previous work of De Luca et al.[32]
| Conclusion | | |
Along with force steadiness improvement, higher visuospatial resolution adds to force fluctuation complexity that indexes abundant force-scaling strategy during static force tracking. The change in force fluctuation structure corresponds with increases in discharge variability across MUs and complexity of global discharge rate due to increased variations in discharge commonality. Despite the modulation of discharge variability, force steadiness is secured by low-pass filter property of the musculotendon complex, in support of a force–discharge decoupling in the high visual condition.
Financial support and sponsorship
This research was partly supported by grants from the Ministry of Science and Technology, Taiwan, R.O.C., under Grant Nos. MOST 105-2410-H-040-009 and MOST 104-2314-B-006-016-MY3.
Conflicts of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
| References | | |
| 1. | Sosnoff JJ, Valantine AD, Newell KM. Independence between the amount and structure of variability at low force levels. Neurosci Lett 2006;392:165-9.  |
| 2. | Vaillancourt DE, Haibach PS, Newell KM. Visual angle is the critical variable mediating gain-related effects in manual control. Exp Brain Res 2006;173:742-50. |
| 3. | Rosenbluth D, Allman JM. The effect of gaze angle and fixation distance on the responses of neurons in V1, V2, and V4. Neuron 2002;33:143-9. |
| 4. | Baweja HS, Patel BK, Martinkewiz JD, Vu J, Christou EA. Removal of visual feedback alters muscle activity and reduces force variability during constant isometric contractions. Exp Brain Res 2009;197:35-47. |
| 5. | Lee Hong S, Newell KM. Visual information gain and the regulation of constant force levels. Exp Brain Res 2008;189:61-9. |
| 6. | Kuznetsov NA, Riley MA. Spatial resolution of visual feedback affects variability and structure of isometric force. Neurosci Lett 2010;470:121-5. |
| 7. | Prodoehl J, Vaillancourt DE. Effects of visual gain on force control at the elbow and ankle. Exp Brain Res 2010;200:67-79. |
| 8. | Chen YC, Lin YT, Huang CT, Shih CL, Yang ZR, Hwang IS. Trajectory adjustments underlying task-specific intermittent force behaviors and muscular rhythms. PLoS One 2013;8:e74273. |
| 9. | Slifkin AB, Vaillancourt DE, Newell KM. Intermittency in the control of continuous force production. J Neurophysiol 2000;84:1708-18. |
| 10. | Vaillancourt DE, Larsson L, Newell KM. Time-dependent structure in the discharge rate of human motor units. Clin Neurophysiol 2002;113:1325-38. |
| 11. | Miall RC, Weir DJ, Stein JF. Manual tracking of visual targets by trained monkeys. Behav Brain Res 1986;20:185-201. |
| 12. | Pasalar S, Roitman AV, Ebner TJ. Effects of speeds and force fields on submovements during circular manual tracking in humans. Exp Brain Res 2005;163:214-25. |
| 13. | Coombes SA, Corcos DM, Sprute L, Vaillancourt DE. Selective regions of the visuomotor system are related to gain-induced changes in force error. J Neurophysiol 2010;103:2114-23. |
| 14. | Poon C, Coombes SA, Corcos DM, Christou EA, Vaillancourt DE. Transient shifts in frontal and parietal circuits scale with enhanced visual feedback and changes in force variability and error. J Neurophysiol 2013;109:2205-15. |
| 15. | Jordan K, Jesunathadas M, Sarchet DM, Enoka RM. Long-range correlations in motor unit discharge times at low forces are modulated by visual gain and age. Exp Physiol 2013;98:546-55. |
| 16. | Schmied A, Pagni S, Sturm H, Vedel JP. Selective enhancement of motor neuron short-term synchrony during an attention-demanding task. Exp Brain Res 2000;133:377-90. |
| 17. | Laine CM, Yavuz SU, Farina D. Task-related changes in sensorimotor integration influence the common synaptic input to motor neurons. Acta Physiol (Oxf) 2014;211:229-39. |
| 18. | Hu X, Rymer WZ, Suresh NL. Assessment of validity of a high-yield surface electromyogram decomposition. J Neuroeng Rehabil 2013;10:99. |
| 19. | Nawab SH, Chang SS, De Luca CJ. High-yield decomposition of surface EMG signals. Clin Neurophysiol 2010;121:1602-15. |
| 20. | De Luca CJ, Adam A, Wotiz R, Gilmore LD, Nawab SH. Decomposition of surface EMG signals. J Neurophysiol 2006;96:1646-57. |
| 21. | Hu X, Rymer WZ, Suresh NL. Accuracy assessment of a surface electromyogram decomposition system in human first dorsal interosseus muscle. J Neural Eng 2014;11:026007. |
| 22. | De Luca CJ, Nawab SH, Kline JC. Clarification of methods used to validate surface EMG decomposition algorithms as described by Farina. (2014). J Appl Physiol (1985) 2015;118:1084. |
| 23. | Nawab SH, Wotiz R, De Luca CJ. Resolving EMG Pulse Superpositions Via Utility Maximization. Proceedings 8 th World Multiconf. Vol. XII. Orlando, FL: Systemics, Cybernetics, Informatics; 2004. p. 233-6. |
| 24. | Lin YT, Kuo CH, Hwang IS. Fatigue effect on low-frequency force fluctuations and muscular oscillations during rhythmic isometric contraction. PLoS One 2014;9:e85578. |
| 25. | Manjarrez E, Rojas-Piloni JG, Méndez I, Martínez L, Vélez D, Vázquez D, et al. Internal stochastic resonance in the coherence between spinal and cortical neuronal ensembles in the cat. Neurosci Lett 2002;326:93-6. |
| 26. | Richman JS, Moorman JR. Physiological time-series analysis using approximate entropy and sample entropy. Am J Physiol Heart Circ Physiol 2000;278:H2039-49. |
| 27. | De Luca CJ, LeFever RS, McCue MP, Xenakis AP. Control scheme governing concurrently active human motor units during voluntary contractions. J Physiol 1982;329:129-42. |
| 28. | Davies RM, Gerstein GL, Baker SN. Measurement of time-dependent changes in the irregularity of neural spiking. J Neurophysiol 2006;96:906-18. |
| 29. | Witham CL, Baker SN. Network oscillations and intrinsic spiking rhythmicity do not covary in monkey sensorimotor areas. J Physiol 2007;580:801-14. |
| 30. | Negro F, Farina D. Factors influencing the estimates of correlation between motor unit activities in humans. PLoS One 2012;7:e44894. |
| 31. | De Luca CJ, Erim Z. Common drive of motor units in regulation of muscle force. Trends Neurosci 1994;17:299-305. |
| 32. | De Luca CJ, Gonzalez-Cueto JA, Bonato P, Adam A. Motor unit recruitment and proprioceptive feedback decrease the common drive. J Neurophysiol 2009;101:1620-8. |
| 33. | Farina D, Negro F. Common synaptic input to motor neurons, motor unit synchronization, and force control. Exerc Sport Sci Rev 2015;43:23-33. |
| 34. | Farina D, Negro F, Muceli S, Enoka RM. Principles of motor unit physiology evolve with advances in technology. Physiology (Bethesda) 2016;31:83-94. |
| 35. | Costa M, Goldberger AL, Peng CK. Multiscale entropy analysis of complex physiologic time series. Phys Rev Lett 2002;89:068102. |
| 36. | Hwang IS, Lin YT, Huang WM, Yang ZR, Hu CL, Chen YC. Alterations in neural control of constant isometric contraction with the size of error feedback. PLoS One 2017;12:e0170824. |
| 37. | Simes RJ. An improvement of Bonferroni procedure for multiple tests of significance. Biometrika 1986;73:751-4. |
| 38. | Fox EJ, Baweja HS, Kim C, Kennedy DM, Vaillancourt DE, Christou EA. Modulation of force below 1 Hz: Age-associated differences and the effect of magnified visual feedback. PLoS One 2013;8:e55970. |
| 39. | Contessa P, Adam A, De Luca CJ. Motor unit control and force fluctuation during fatigue. J Appl Physiol (1985) 2009;107:235-43. |
| 40. | Taylor AM, Christou EA, Enoka RM. Multiple features of motor-unit activity influence force fluctuations during isometric contractions. J Neurophysiol 2003;90:1350-61. |
| 41. | Yao W, Fuglevand RJ, Enoka RM. Motor-unit synchronization increases EMG amplitude and decreases force steadiness of simulated contractions. J Neurophysiol 2000;83:441-52. |
| 42. | Hawkins KM, Sayegh P, Yan X, Crawford JD, Sergio LE. Neural activity in superior parietal cortex during rule-based visual-motor transformations. J Cogn Neurosci 2013;25:436-54. |
| 43. | Nafati G, Rossi-Durand C, Schmied A. Proprioceptive control of human wrist extensor motor units during an attention-demanding task. Brain Res 2004;1018:208-20. |
| 44. | Stacey WC, Lazarewicz MT, Litt B. Synaptic noise and physiological coupling generate high-frequency oscillations in a hippocampal computational model. J Neurophysiol 2009;102:2342-57. |
| 45. | Ward LM, MacLean SE, Kirschner A. Stochastic resonance modulates neural synchronization within and between cortical sources. PLoS One 2010;5:e14371. |
| 46. | Baldissera F, Cavallari P, Cerri G. Motoneuronal pre-compensation for the low-pass filter characteristics of muscle. A quantitative appraisal in cat muscle units. J Physiol 1998;511 (Pt 2):611-27. |
| 47. | Negro F, Muceli S, Castronovo AM, Holobar A, Farina D. Multi-channel intramuscular and surface EMG decomposition by convolutive blind source separation. J Neural Eng 2016;13:026027. |
| 48. | Zhu X, Zhang Y. High-density surface EMG decomposition based on a convolutive blind source separation approach. Conf Proc IEEE Eng Med Biol Soc 2012;2012:609-12. |
| 49. | Farina D, Enoka RM. Surface EMG decomposition requires an appropriate validation. J Neurophysiol 2011;105:981-2. |
| 50. | Chen YC, Lin YT, Chang GC, Hwang IS. Paradigm shifts in voluntary force control and motor unit behaviors with the manipulated size of visual error perception. Front Physiol 2017;8:140. |
| 51. | Farina D, Holobar A, Merletti R, Enoka RM. Decoding the neural drive to muscles from the surface electromyogram. Clin Neurophysiol 2010;121:1616-23. |
| 52. | Adam A, De Luca CJ, Erim Z. Hand dominance and motor unit firing behavior. J Neurophysiol 1998;80:1373-82. |
| 53. | Moritz CT, Barry BK, Pascoe MA, Enoka RM. Discharge rate variability influences the variation in force fluctuations across the working range of a hand muscle. J Neurophysiol 2005;93:2449-59. |
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3]
|
Search Pubmed for