Experiments on the resolution of electromyographic signals in muscle fibers
Experiments on the resolution of electromyographic signals in muscle fibers
The precise EMG analysis techniques covered in this chapter recover all available information in the EMG signal.EMG signal analysis consists of two categories: morphological analysis and control characterization. Morphological analysis refers to the description of MUAP waveform parameters such as peak-topeakamplitude, timeduration, numberofphase, and area, which will be obtained by recovering the MUAP.
Operation method
Experiments on the resolution of electromyographic signals in muscle fibers
Materials and Instruments
Four-lead needle electrodes Computerized high-speed data acquisition boards Move move I. Signal Detection An important feature of the Precision Resolution Technique signal detection is the use of a special electrode to detect three EMGff signals in the muscle tissue, which is a relatively basic and important aspect of the technique. The three signals are necessary to minimize doubt in the decision-making process, and therefore the different MUAPTs in the EMG signals are identified according to computer algorithms.This recording is accomplished using specially designed QuacJrJfilarectrode electrodes, which consist of two types of electrodes: a needle electrode, and a wire electrode. （The electrodes consist of two types: a needle electrode and a wire electrode.) Needle electrodes have the following advantages: I can be used to obtain morphological characteristics of action potentials It can be used to obtain macro-EMG signals Can be precisely localized in muscle -Inserts electrodes into high quality signal areas by hand manipulation Can be used in the clinic -It cannot be used for recording of dynamic contractions They can cause pain when inserted into deep muscle tissue. Cannot be left in the muscle tissue for long periods of time, usually less than Ih Metal wire electrodes have the following advantages: They can be used to record slow dynamic contractions. painless Can be left in muscle tissue for long periods of time, typically up to several hours They cannot be repositioned, making it more difficult to probe sites that produce signals that are more amenable to resolution. they can wander during muscle contraction Figure 27-3A shows a detail of a needle electrode and the elements connected to the amplifier. The distinctive feature of this electrode is that there are four probe surfaces on the sidewall of the metal tip, with guidewires of 50 um or 75 um diameter depending on the density of the fiber distribution in the muscle under study, each spaced at a corresponding distance of 150 um or 200 um from each other.These configurations were experimentally tested and are perfectly suited to the quality of the signal required for the analytical operations. The electrode probe surfaces were connected in a differential amplifier bipolar configuration with three different EMG signal outputs. We found that the smaller the electrode probing surface the better the selectivity, which is suitable for muscles with a high density of myofiber distribution and for forceful contractions. In addition, this electrode can probe both concentric electrode EMG signals and macro-EMG signals by another configuration. Figure 27-4 shows a specific example of use. Concentric EMG signals can be probed through a plug with a needle tip that has the same probing surface area as a standard concentric electrode plug (2.7 mm2 ), while the probing surface for macro-EMG signals is standard. Therefore, with this configuration, the four-wire needle electrode can probe three EMG signals from the side wall of the tip, and because there are a total of four inputs, it is also possible to probe concentric electrode EMG signals or macro-EMG signals at the same time. Figure 27-3B shows the metal wire electrodes with the connected amplifier components. The electrodes are made of four nickel-cadmium alloy wires or platinum wires wrapped in a nylon sheath, with wire diameters of 50 um or 75 um depending on the probing selectivity, the finer the wire diameter the better the selectivity, and insulated wires that have a small area of exposed wire ends that are curved to form fishhooks at the end of the wire at 1 mm to facilitate anchoring in the muscle. The four-guide needle electrodes that can be used to characterize multiple motor units have high selectivity, but add some difficulty in acquiring stable EMG signals. For example, a slight movement may have no effect on signals recorded with concentric circle electrodes or coarse needle electrodes, but it has a significant effect on signals recorded with four-guide needle electrodes. Therefore, we have invented some technical methods for inserting and positioning such electrodes that can minimize the movement and enhance the signal-to-noise ratio. The most stable EMG signals are obtained when the electrode is inserted obliquely into the muscle fiber at an angle of 30 degrees. This is because if the electrode is inserted parallel to the muscle fiber, then the muscle fiber is likely to slide against the electrode during diastole; conversely the greater the angle of the electrode to the muscle fiber ^ the greater the shear force at the tip of the electrode, which also tends to cause the electrode to slide. Therefore it is important to find as much as possible an angle at which the electrode tip shear force is sufficient to anchor the electrode to the sliding muscle fibers in order to minimize electrode movement, and this is the basis for the 30-degree angle we propose. If possible, the detection surface of the electrode can be positioned at the point of movement, where the recorded MUAP propagates in the opposite direction, the polarity of the MUAP is reversed when recorded with a four-guide needle electrode, and the opposite waveform signals are more favorable for identification and resolution. Due to the unique design mentioned above, the four-guide needle electrode requires a special insertion and positioning technique, unlike the standard concentric circle electrode. Concentric electrodes have a relatively thick and pointed detection surface, so rotating the electrodes makes no significant difference to the signal quality compared to moving the prongs in and out. Concentric round electrodes can therefore be inserted and retracted straight, or the needle can be fed multiple times. However, the detection surface of the four-guide needle electrode is small and located on the side wall of the tip, so the electrode is very sensitive to rotational needle feed. In view of this, when positioning the four-guide needle electrode, the electrode is rotated forward and backward while slowly inserting and retracting, which allows the experimental operator to greatly reduce the number of times the needle is inserted, and at the same time increases the effective area of the detection site. Listening to the sound is very useful in the detection of MUAP. When the detection surface of the four-guide needle electrode is close to an active muscle fiber, the amplitude of the signal and the frequency of the discharge gradually increase, and if an amplified monitor is connected, a "crackling" sound with a higher and higher pitch will be heard. By listening to the feedback, the operator can concentrate on positioning the electrodes without the need to frequently look back at the monitor or oscilloscope, which facilitates the rapid separation of high-quality signals. It generally takes a few minutes to position the electrode plate at a motor unit with a well-defined waveform, and if the subject's muscles maintain a high level of contraction during the electrode positioning process, then muscle fatigue will soon set in. However, if the contraction of the muscle is low, it is difficult for the electrode to record an accurate signal on the muscle it is about to be anchored to. Therefore, when positioning the electrodes, the subject's muscle contraction force should be maintained at 10% of the maximal arbitrary contraction force, and then 2-3 significant motor unit waveforms should be isolated. Once a suitable position is found, the subject can be allowed to slowly increase the muscle contraction force until the pre-determined number of MUAPTs is obtained under monitoring, which generally requires 2-3 motor units to be observed. If the recorded signal is unstable during increased muscle contraction force, the search can be abandoned and resumed. The greatest difficulty with electrode positioning is how to stabilize the recorded signal when changing muscle contraction force. Even during isometric contraction, muscle fibers may move slightly relative to the fascia, skin, and electrodes, and this slight movement may cause the signal to become unstable. However, there are some solutions to these problems, such as paying attention to how the electrodes move during muscle contraction and how the signal changes, and then positioning the electrodes so that the movement itself not only does not affect the recording, but also improves the quality of the signal If this is also unsuccessful, then you can also try to pinch the electrodes to make them stay still, which is useful, but also problematic. This method is useful but also problematic, because the experimenter wants to fix the electrode's detection surface relatively to the local muscle fiber, so it is inevitable that the electrode will need to be moved in order to fix the electrode's detection surface relatively steadily to the recorded muscle fiber. Pinched electrode immobilization is likely to limit the achievement of the above purpose, so it is important to remember that the high-fidelity signal can only be determined by the position of the electrode in relation to the muscle fiber, not by the position of the electrode held in the hand. Traditionally the first step towards achieving the recording of EMG signals is to amplify the signal as much as possible to avoid distorting the signal before it is digitized, and for this purpose the maximum sampling spectrum is applied in order to improve the sampling resolution of the digitized signal. Therefore, the conventional method suggests adjusting the amplifier gain to the maximum to ensure that the EMG signal waveforms are free of tangents. However, the sampling spectrum of the signal recorded with four-guide needle electrodes or four-guide wire electrodes is 1 to 10 kHz, which is a unique feature in the precise resolution procedure. Precision parsing techniques purposely distort the action potential waveform to produce signal shapes unfamiliar to the researcher, which are required for parsing algorithms. Because the shorter and sharper the wave shape is, and because the length of the MUAP tail decreases, the less chance it has of overlapping in multiple MUAPs. Applying this parsing technique, concentric electrode EMG signals are sampled from 10 Hz to 10 kHz, while macro EMG signals are sampled from 10 Hz to lkHz. When the signal of any channel exceeds a pre-set threshold, it can be sampled and digitized. The sampling threshold is a default value suitable for a certain typical signal, and if the threshold is set appropriately, then it is likely to improve the signal-to-noise ratio of the signal being sampled and analyzed. However, this operation requires specific practice and the operator should be familiar with the program software of the system. Because only signals on the threshold can be captured and stored, setting the sampling threshold can greatly improve the efficiency of data storage. The system can provide a complete set of time references to store the amount of time of the acquired signals, i.e., there is a point in time between acquired signals that represents a signal loss, so that all signals to be parsed are not missed. Figure 27-5 shows an operator-assisted parsing process, with the screen view showing the time-compressed data. The number below the vertical line represents the length of signal leakage (expressed in ms). The signals recorded on the sidewall of the tip of the four-guide needle electrode are called microEMG, and these signals are emitted from approximately 3-4 muscle fibers per motor unit. Therefore, microEMG is not as selective as the signals recorded with a single-fiber electrode, but it is much more selective than the signals recorded with a concentric circle electrode. The micro-EMG was acquired at a sampling frequency of 50 parity, which is just above the Nyquist frequency (see Chapter 45). High-resolution MUAPs are important in comparative analytical operations because an important feature of precise analytical techniques is to resolve the decision space in the time domain. Because the concentric electrode signals and macro-EMG signals have a narrow frequency iP, the sampling frequency is low, typically 2 kHz. The parsing procedure is implemented on the basis of complex rules of algorithms, which have been created over a period of 20 years, and which contain all the information on how to deal with the know-how of actually testing the properties of EMG data. These algorithms identify action potentials by using templatematching and probabilityoffiringstatistics, resolve superimpositions and assign action potentials to motor units. Accuracy and modification can be checked by using user-interactive editing algorithms based on established rules, see DeLuca (1993) for details. By applying the temporal record of MUAP discharge in each MUAPT created by parsing tiny EMG signals, it is also possible to extract the MUAPs of concentric electrodes and macro-EMGs from the corresponding EMG signals.This process can be accomplished by either waveformaveraging or the physiological term peak potential-triggeredaveraging (spike- triggeredaveraging) (see Chapter 18). The method is that when a MUAP of a particular motor unit occurs, the time interval corresponding to the signal can be selected and saved, and then the time sum is averaged. In this process, the time intervals belonging to the portion of the waveform that is a MUAP must be totaled, and the portion that is not related to the MUAP must be discarded because the positive and negative phases of action potentials from other sources may overlap with it. This averaging method is particularly suitable for long time intervals, noise, few signals, low amplitude and few synchronized discharges. Another factor that determines the effectiveness of the trigger averaging method is that the processable MUAP must be displayed on both tiny and concentric signals. The geometry of the four-conducting electrodes is constructed with this purpose in mind. The processed MUAP of concentric circle electrodes and macro-EMG is shown in Figure 27-4. When an action potential is recognized as belonging to the activity of a particular motor unit, the maximum magnitude can be obtained by an algorithm and the time of its occurrence can be saved, thus obtaining all the discharge timescales of the corresponding unit. Figure 27_6A shows the impulse release interval versus contraction time. Dot plots are often used to check for resolution errors during user interaction editing. To illustrate motor unit activity timing, Figure 27-6B shows a line plot of action potential release times. Individual discharges help in the characterization of motor unit synchronization, and some other discharging-discharging relationships, such as reflexes, can also be studied. However, the method provides less information on how motor unit discharges are modulated, and thus it is of interest to further investigate the disbursement frequency characteristics from which more of their mechanical relevance can be obtained. There are a number of methods to obtain the firing frequency. We prefer to use Hanningwindow's low-pass filtering function to obtain signals with the mean values of release duration and discharge frequency for each motor unit impulse string, and the most commonly used width is 400ms, however, in practice it should also depend on the specifics of the extracted signals. Figure 27-6C shows the average value of the discharge frequency of the motor unit at 400ms. The application of this method in neuroscience focuses on the characterization of the firing behavior of activated motor units, and therefore we will describe the results. For a clinical application of the exact resolution technique see DeLuca (1993). Precision resolving analysis was first used to observe discharge frequency decay (DeLucaandForrest1973;DeLuca1985;DeLucaetal.19%); we reported a decrease in the frequency of motor unit discharges over time during isotonic and isovolumic contractions (Fig. 27-7A), and no new motor unit discharges were seen during the first 20 s of contraction. We first suggested (DeLuca 1979) and further revealed (DeLuca 1996) that the pattern of decreasing discharge frequency during sustained spontaneous contraction can account for two phenomena, namely (i) the intrinsic property of individual motor neuron discharge frequency to decay over time, first reported by Kemell (1965), who used direct current to stimulate animals; and (ii) a simplified method of activating motor units is to increase the amplitude and duration of muscle tone on the basis of repetitive discharges, which is often referred to as twitchpotentiation. The second phenomenon we have observed is co-driving (DeLucaetal.1982a,b), whereby fluctuations in discharge frequency between motor units do not actually have a time delay. This is illustrated by the reciprocal correlation of the discharge frequencies of multiple units activated at the same time (Fig. 27-7B). This phenomenon was seen in all the muscles tested, including both the fine terminal muscles and the coarse proximal muscles. Even motor units belonging to different motoneuron innervations, when fluctuations were controlled within the range of one functional unit, other units showed a common pattern of fluctuations, as seen in our antagonist muscle coactivation test (DeLucaandMambrito1987). This phenomenon has also been confirmed by other scholars (Miles1987;StashukanddeBruin1988;Guiheneuc1992;IyeretaL1994;Semmleretal.1997).This suggests that the CNS has evolved a relatively simple set of strategies for regulating motor unit activity. In addition, the irregular nature of the discharge frequencies and the co-driving phenomenon that we found suggests that muscles are unlikely to produce smooth and sustained dystonia. In this regard, we verified this by experiments on the relationship between the reciprocal correlation of discharge frequencies and muscle tone. The tests showed a significant correlation between the latency of the mechanical delay in the formation of muscle fiber tension and the transmission of muscle tension in muscle and tendon tissue. The high degree of intercorrelation of the frequency of impulse delivery between motor units does not mean that the activity of each unit is synchronized. By synchronization, we mean that the motor units emit to each other at a certain fixed time latency, which exists in two forms: long and short duration (Fig. 27-8). An isometric, isotonic contraction test of six paired motor units showed that 8% of the discharges were short-duration synchronized, whereas 1% were long-duration synchronized (DeLucaetaL1993). Short-duration synchronization occurs sporadically at certain time intervals and can also be seen in clusters of one or two consecutive discharges that typically have no discernible effect on muscle tone production. We consider motor unit discharge synchronization to be an anomaly not designed for its own physiological function. The third is the onionskinphenomenon. With Person and Kudina (1972) and Tan-ji and Kato (1973)— we first reported that the average discharge frequency of motor units activated first was higher than that of those activated later during isometric contractions lasting less than 20 s. (The average discharge frequency of motor units activated first was higher than that of those activated later during isometric contractions. (Examination of the temporal distribution of motor unit discharge frequencies reveals a hierarchical distribution of amplitudes overlapping each other, similar to the appearance of onion skin. See Figure 27-6C.) Subsequently, we reported this phenomenon in detail (DeLucaetal.1982a)〇 Hoffer et al. (1987), and Stashuk and deBruin (1988) confirmed it, respectively. The fast contractile units that are recruited later are easily activated by glycolysis and the slow contractile units that are recruited first are easily activated by aerobic oxidation, and the former discharge at a higher frequency than the latter but are less likely to elicit a contractile response. Even at high levels of contraction near 80% to 90% MVC, the discharge frequency of the high-threshold motor units only reached 20?30 times per second, which was insufficient to elicit a fully tonic contraction. This result is contrary to the previously held view that the higher the discharge frequency, the greater the tonicity produced. This raises the question: why is the motor unit so modulated that it prevents the muscle from reaching its maximum potential for myotonic force production? After all, in order to generate myotonia, the regulatory mechanism should be fully mobilizing every possible pathway. Why has the muscle evolved an apparent reservecapacity during voluntary contraction, and why is this not readily activated? This is an extremely interesting question. One possible explanation is that high threshold motor units, which are prone to rapid fatigue, fatigue quickly if discharged rapidly, but are not subjected to sustained muscle contraction at high tension levels through the action of a regulatory system, which is necessary to deal with critical situations and to ensure survival. It is clear that the modulatory action of the motor unit allows an optimal combination of muscle tone and contraction time, rather than just a unilateral regulation of muscle tone. It has been reported that muscles can produce certain abnormal properties during critical situations, and this partially reserved function may play a role here. The behavior of discharge frequency also exhibits two spontaneous phenomena, namely (i) postrecruited motor units have a higher initial discharge frequency, which is consistent with the phenomenon described by Clamann (1970), and (ii) when maximal contraction is reached, the discharge frequency of all motor units converges to a common value (DeLucaandErim1994;ErimetaL1996). The above results suggest that the control (net excitatory) signal acts as a unit on the motor neuron pool. This is as Hermeman and colleagues have argued, that the respective properties of the motoneurons determine their level of recruitment in response to net excitation (Hennemanetal.1965a,b). We add to their illuminating observation that the firing frequencies of individual motor units in response to net excitation are simultaneous and synchronized, and that the average firing frequency is also distributed according to a certain hierarchy and is negatively correlated with the recruitment threshold. The fourth phenomenon observed is regulatory diversity (DeLucaetal.1982a). Small terminal muscle motor units, such as the first dorsal interosseous muscle, begin to recruit at 50% MVC and the mean discharge frequency reaches a high value at 80% MVC (approximately 40 PPS); whereas those of large proximal muscles, such as the deltoid and trapezius, whose motor units begin to recruit at 80% MVC, have a relatively low discharge frequency (approximately 30 pPS). A similar phenomenon was observed by Kukulka and Qamann (1981) in the wrestling adductor and limb biceps. The reduction in the power range of large proximal muscles is most likely the result of enhanced repetitive inhibition of the Renshaw system, as the effect of this system on these muscles is significant, see Rossi and Mazzachio (1991). The diverse regulatory properties are of interest in at least two ways. First, it favors the production of smooth muscle forces; small muscles have fewer motor units than large muscles that have more (metric or otherwise) motor units, so that the increase in force caused by unit recruitment appears somewhat coarser across the range. Second, large proximal muscles tend to be postural maintenance, requiring frequent sustained contractions, and the lower discharge frequency helps to delay fatigue. We have found that prolonged exercise induces modifications in the regulatory properties of motor units (Adametal.1988). By comparing the regulatory parameters of motor units during isometric and isotonic contractions of the first dorsal interosseous muscle of the dominant and nondominant hand at the same MVC7jC level, we found that: reaching the same level of contraction, the discharge frequency was lower on the side of the dominant hand, and the lower the level of tension, the more motor units were recruited. This result is consistent with the previously known fact that the dominant hand has slow-contracting fibers, most likely due to long-term exercise. At lower discharge frequencies, the fusion of slow contractile fibers can maintain the excitatory state without causing diminished tone in the dominant hand. Recently, we reported that age factors can cause changes in the regulatory properties of motor units (Erimetal. in press). In our study, we found that when subjects were older than 65 years, the firing frequency and recruitment threshold of the motor units of the first dorsal interosseous muscle were modulated in the same way as after functional exercise. This phenomenon is not surprising, since it is known that type I bradykinesia fibers are higher in aged muscles and that they are also elevated in exercised muscles, although not for the same reasons. In the course of co-driving studies in aging subjects, we found that about half of the subjects showed a marked reduction in the correlation of paired motor units, or even a complete disappearance in some individuals. At the same time, the onion skin phenomenon appeared to be disturbed (Fig. 27-9). By examining the temporal distribution of discharges, we found that many motor unit discharges overlapped with those of earlier recruitments, and the discharge behavior was no longer hierarchical and orderly, with some enhanced and some diminished during isometric and isotonic contractions. We hypothesized that incoordination of discharge frequencies between motor units would produce ineffective dystonia. All observations were made during isometric and isotonic contractions of fairly short duration (less than 20 s). These results may not fully characterize the behavior of those limb-end and postural-maintenance muscles that require frequent and sustained contractions for their regulatory properties. Recently, in a study of normal healthy deltoid short-duration (20s or less) contractions and normal healthy first dorsal interosseous muscle long-duration (150s or more) contractions, we observed disturbances in the onion skin phenomenon. We hypothesize that the crossover in discharge frequency is related to at least two factors that can cause earlier-recruited motor units to discharge less frequently than newer recruits: 1) Repeated inhibition of the first-recruited motor units by the Renshaw system. This system has a significant effect on proximal muscles, such as the deltoid muscle, and, as a result, onion-skin phenomena can be disturbed during short-duration contractions. (ii) Kemell (1965) reported the process of motor neuron adaptation, in which the continuous activation of motor units causes a decrease in the firing frequency, thus causing a lower firing frequency in the first recruited motor units than in the later recruited ones. We clearly observed the phenomenon of motor unit k substitution (WestgaardandDeLuca1999) during long-duration contractions of 5-60 min studied together with Westgaard. After a sustained contraction, the motor unit activity level was mildly reduced and discharge ceased; and when the subsequent tension output was mildly increased, a newly recruited unit appeared at the site of de-recruitment (Fig. 27-10). We believe that this is a result of motor neuron recruitment threshold adaptation. If a motor unit has been activated for some time then it is recruited at a higher threshold than the immediate follower. It is in this manner that neighboring motor units are recruited in response to a net increase in excitability of the motor neuron pool. For more product details, please visit Aladdin Scientific website.
Needle electrodes have the following disadvantages:
Metal wire electrodes have the following disadvantages:
