An engineering approach to improve clinical usability of quantitative electromyography

Electromyography (EMG) is the record of the electrical activity from muscle fiber membranes. This invaluable clinical tool in neurology aids in the diagnosis and monitoring of disease affecting muscle and nerve. Routine clinical EMG studies rely on the experience of the physician to analyze the data in a qualitative manner. Quantitative EMG (QEMG) refers to a number of techniques that measure various aspects of the EMG signal and result in statistical data. These techniques are becoming broader in scope, more automated and increasingly available on EMG machines. However, QEMG studies have challenges in a number of operational parameters from the engineering perspective but may not accurately fit from the physiologic and pathologic perspectives, and a number of these issues have not been investigated in a systematic way. Here we present a number of studies aimed at validating and improving clinical usability of QEMG. First, we compare three QEMG algorithms available on EMG machines for use in the clinic, a study not performed previously. We determined that two algorithms yield similar results with minimal user intervention, while the third requires considerable expert review of the results and which are less robust than with the first two. Second, we show that among available sizes of intramuscular needle electrodes the smaller diameter electrode yields data comparable to the larger diameter electrode for clinical QEMG. Third, we show that any needle electrode position along the longitudinal axis of the muscle with respect to the distribution of neuromuscular junctions within the muscle is acceptable for clinical QEMG studies. Fourth, we investigate and find that high-pass filtering is not an effective means of extracting more sensitive information from the EMG signals. Finally, we determine that at each position of the electrode within the muscle, 10 s worth of data collection balances the need to collect sufficient data with the possibility of degrading the signal due to subtle physiologic movements. The results of these efforts are a better understanding of the practical limits of the QEMG algorithms and how operational parameters can be optimized for more accurate statistics, more rapid data acquisition, and greater patient tolerability.

AN ENGINEERING APPROACH TO IMPROVE CLINICAL USABILITY OF QUANTITATIVE ELECTROMYOGRAPHY by Alexander Arthur Brownell A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Bioengineering The University of Utah May 2011 Copyright © Alexander Arthur Brownell 2011 All Rights Reserved The Uni v e r s i t y of Ut a h Gr ad ua t e Sc ho o l STATEMENT OF DISSERTATION APPROVAL The dissertation of Alexander Arthur Brownell has been approved by the following supervisory committee members: Mark B. Bromberg , Chair 17 Dec 10 Date Approved Richard A. Normann , Member 17 Dec 10 Date Approved Douglas A. Christensen , Member 17 Dec 10 Date Approved Rob S. MacLeod , Member 21 Dec 10 Date Approved J. Robinson Singleton , Member 4 Jan 11 Date Approved and by Richard D. Rabbitt , Chair of the Department of Bioengineering and by Charles A. Wight, Dean of The Graduate School. iii ABSTRACT Electromyography (EMG) is the record of the electrical activity from muscle fiber membranes. This invaluable clinical tool in neurology aids in the diagnosis and monitoring of disease affecting muscle and nerve. Routine clinical EMG studies rely on the experience of the physician to analyze the data in a qualitative manner. Quantitative EMG (QEMG) refers to a number of techniques that measure various aspects of the EMG signal and result in statistical data. These techniques are becoming broader in scope, more automated and increasingly available on EMG machines. However, QEMG studies have challenges in a number of operational parameters from the engineering perspective but may not accurately fit from the physiologic and pathologic perspectives, and a number of these issues have not been investigated in a systematic way. Here we present a number of studies aimed at validating and improving clinical usability of QEMG. First, we compare three QEMG algorithms available on EMG machines for use in the clinic, a study not performed previously. We determined that two algorithms yield similar results with minimal user intervention, while the third requires considerable expert review of the results and which are less robust than with the first two. Second, we show that among available sizes of intramuscular needle electrodes the smaller diameter electrode yields data comparable to the larger diameter electrode for clinical QEMG. Third, we show that any needle electrode position along the longitudinal iv axis of the muscle with respect to the distribution of neuromuscular junctions within the muscle is acceptable for clinical QEMG studies. Fourth, we investigate and find that high-pass filtering is not an effective means of extracting more sensitive information from the EMG signals. Finally, we determine that at each position of the electrode within the muscle, 10 s worth of data collection balances the need to collect sufficient data with the possibility of degrading the signal due to subtle physiologic movements. The results of these efforts are a better understanding of the practical limits of the QEMG algorithms and how operational parameters can be optimized for more accurate statistics, more rapid data acquisition, and greater patient tolerability. To my loving wife Michelle, whose long-suffering made this possible. Also, to Mark, a wonderful friend and unparalleled mentor. TABLE OF CONTENTS ABSTRACT .................................................................................................................. iii 1 INTRODUCTION ..................................................................................................... 1 Anatomy and Physiology of the Motor Unit ................................................... 2 Motor Unit Action Potential ........................................................................... 5 EMG Studies................................................................................................ 24 2 RATIONALE FOR THE WORK AND TECHNIQUES EMPLOYED .................... 32 Algorithm Comparison ................................................................................. 33 High-pass Filtering ....................................................................................... 46 Temporal Dispersion of the MUAP .............................................................. 54 Needle Electrode Selection........................................................................... 64 Optimizing Acquisition Time ....................................................................... 71 3 CONCLUSIONS AND DISCUSSION .................................................................... 79 Future Work ................................................................................................. 86 REFERENCES.............................................................................................................. 91 1 1 INTRODUCTION Electromyography (EMG) is the record of electrical activity generated by muscle fiber membranes measured by electrodes placed either within or external to muscle. The membrane depolarization that is measured is the signal that initiates muscle fiber contraction (excitation-contraction coupling). EMG techniques provide information directly about the function and pathology of muscles, and indirectly about the function and pathology of both the nerves that innervate them and neuromuscular junction transmission. Thus, EMG is a useful tool to determine if a muscle is normal or abnormal; and if abnormal, whether the primary disease process is of nerves innervating the muscle or of the muscle itself, and also the integrity and function of the neuromuscular junction. Simply stated, EMG can estimate the architectural arrangement of muscle fibers within the muscle and their activation. This body of work bridges the two disciplines concerned with the improvement of EMG techniques. Engineers have the ability to create systems and algorithms which allow for the measurement of EMG signals and can create new techniques to improve accuracy and efficiency. Clinicians have an understanding of the anatomy, physiology and disease processes underlying the signals which are generated by the EMG study. We explore clinically relevant questions to refine and optimize data acquisition and 2 interpretation from an engineering perspective with consideration of practical clinical limitations. Anatomy and Physiology of the Motor Unit The motor unit is the basic unit of neuromuscular function, the quantal unit of movement. It is defined as a single α-motor neuron, originating from the anterior horn of the spinal cord, along with its axon, and all of the muscle fibers which it innervates (Figure 1.1). This concept of the motor unit as the final common pathway of movement was introduced in 1925 by Lidell and Sherrington (1). Interposed between the nerve ending and the muscle fiber is the neuromuscular junction. Figure 1.1: Representation of a motor unit. The α-motor neuron at the left is in the anterior horn of the spinal cord. Its axon exits the spinal cord through the anterior root and courses alongside other axons to a muscle. Within the muscle the axon branches up to several hundred times, and each branch innervates a single muscle fiber. 3 The architecture of muscle fibers of a motor unit within muscle has been determined by glycogen depletion studies. Edstrom and Kugelberg, in 1968, found that in rat preparations a single α-motor neuron (ventral rootlet) can be isolated and repeatedly stimulated thus depleting the glycogen stores in all muscle fibers innervated by that motor unit. When the muscle was removed and prepared for histology a cross section of the muscle, when stained for glycogen, revealed depleted fibers representing muscle fibers innervated by the stimulated neuron (2). An example of such a study is shown in Figure 1.2. These studies revealed many previously unconfirmed properties of motor units. Many motor units overlap-sharing much of the same territory in the muscle. Motor units cover relatively large areas of the overall axial territory of the muscle but vary greatly in area, but no motor unit in large skeletal muscle covers the entire cross-sectional area of the muscle. Importantly, relatively few neighboring muscle fibers belong to the same motor unit. Additionally, motor units are essentially uniform in Figure 1.2: Glycogen depletion study in the rat showing muscle in cross section. Single motor neuron activated repeatedly until all glycogen stores are depleted. Excision of muscle and staining for glycogen reveals white muscle fibers belonging to stimulated unit. Modified from Edström and Kugelberg (1968). 4 regard to histochemical fiber type-the innervating α-motor neuron is the main factor determining muscle fiber type and its attendant properties. Studies on felines have suggested that a single muscle contains hundreds of motor units, each unit containing an average of 400-800 muscle fibers, and the overlap of such units is such that any area of the muscle may contain 20-50 overlapping units (3,4). Glycogen depletion studies are still the only method of directly observing the arrangement of muscle fibers of a motor unit. However, this is a destructive procedure and cannot be used clinically. EMG studies remain the most widely used tool for determining the characteristics of motor units in the clinical setting. An estimate of the motor unit territory within human muscle and the arrangement of muscle fibers therein were determined by two groups using different EMG techniques. Buchthal et al. showed that the motor unit could be discovered anywhere longitudinally within a muscle using two multielectrodes inserted into the muscle at right angles as at least a portion of the muscle fibers of the motor unit ran completely from one tendon to the other (5). They showed that most motor unit territories are basically circular. Stålberg and Antoni used a single fiber electrode as a trigger source and a concentric needle electrode as a sampling electrode moved through the muscle at defined steps and thus determined the cross-sectional area and fiber distribution within the motor unit (Figure 1.3). It was determined that most motor unit territories fall between 5 to 10 mm in diameter. Both of these studies showed that fibers within the motor unit are not evenly distributed throughout. 5 Figure 1.3: Scanning EMG study of a normal tibialis anterior muscle from Stålberg and Antoni (6). Of note, the resulting electrical potential varies greatly depending on the position of the electrode within the motor unit. The different action potential morphologies result from the spatial relationship of the needle to the nearest muscle fibers with a nonuniform distribution of muscle fibers within the motor unit. Motor Unit Action Potential The MUAP is the electrical view of the motor unit and consists of the combined potentials from all of the single muscle fiber action potentials within the uptake area of the electrode. Surface electrodes have a large uptake area. Intramuscular electrodes vary in size and hence uptake areas and yield a variety of restricted views of the electrical motor unit. In this regard, the electrode may be thought of as a probe and the muscle a 6 black box. This black box is unusually complex with respect to the number of motor units, the size of the motor units, the recruitment order with voluntary activation, and the arrangement of neuromuscular junctions. Further, the black box may be normal or abnormal due either to a neurogenic or myopathic process. In the clinical investigation the black box can be probed with a variety of electrode types whose position in the muscle and relation to fibers of the motor units is unknown to the investigator. We will deal with these issues in greater depth later in this section. There are common metrics used to describe an MUAP waveform which include: peak-to-peak amplitude, duration, area, area-to-amplitude ratio, number of turns, and number of phases (shown in Figure 1.4). The metrics allow for quantitative statistics, and from clinical empiric experience, can help in distinguishing normal from pathologic muscle. Abnormal MUAPs exhibit characteristic features that, in combination with other clinical data, can confirm the diagnosis of a number of pathologic conditions. Initially, in the 1950's, motor units from normal subjects and those with known pathology were studied from photographs of the oscilloscope screen to quantify the range of metric values. It is from these laboriously collected data that qualitative interpretation of normal from abnormal motor units is determined during routine EMG studies. Normal values for these metrics differ among muscles and with age, and a sample recorded with a concentric needle electrode are shown in Table 1.1. The concentric needle, monopolar needle, and single fiber needle are common intramuscular electrodes in clinical use (Figure 1.5). The concentric needle electrode consists of a stainless steel cannula, which acts as the reference electrode, and a central silver wire insulated from the cannula, which is the active electrode. The end of the 7 Figure 1.4: Common metrics obtained from motor unit action potential. Table 1.1: Normal values for quantitative EMG studies recorded with a concentric needle electrode (7). Mean values and standard deviations Amplitude (μV) Duration (ms) Area/Amplitude Phases Turns Muscle Mean SD Mean SD Mean SD Mean SD Mean SD Deltoid 550 ± 110 10 ± 1.3 1.6 ± 0.2 3 ± 0.3 4.2 ± 0.8 Biceps brachii 436 ± 115 9.9 ± 1.4 1.5 ± 0.2 2.6 ± 0.3 4.2 ± 0.6 Dorsal interosseous (FDIH) 752 ± 247 9.4 ± 1.3 1.4 ± 0.2 3.1 ± 0.4 3.9 ± 0.6 Vastus Lateralus 687 ± 239 12 ± 1.9 1.7 ± 0.2 3 ± 0.3 4.5 ± 0.8 Anterior tibialis 666 ± 254 11 ± 1.2 1.7 ± 0.2 3.2 ± 0.3 4.7 ± 0.9 8 Figure 1.5: Representation of three different types of needle electrodes. A: concentric needle electrode-a solid conductive core is the active electrode which is insulated from the needle cannula that is used as the reference electrode. There are two sizes of concentric electrodes. B: monopolar needle electrode-a wire is insulated and the end tapered to a cone which is the active electrode and a separate electrode, often on the skin, is the reference electrode. C: single fiber needle electrode-a 25 μm insulated wire is brought out from the side of the cannula and is the active electrode and the cannula is the reference electrode. D: macroEMG needle electrode-longer single fiber EMG electrode with the distal 15 mm of the cannula uninsulated and used as recording surface. Figure modified from Bromberg 1993 (8). standard sized needle is cut off at 15° to create an elliptical recording surface. There are practical limitations to concentric electrodes based on their stiffness and ability to pass through muscle and limitations to their recording uptake radius, resulting in two diameters for practical electrodes: the routine concentric electrode is a 26 gauge (0.46 mm diameter) needle with a recording ellipse of 580 x 150 μm (0.07 mm2); and the pediatric concentric electrode is a 30 gauge (0.3 mm diameter) needle with a recording ellipse of 390 x 100 μm (0.03mm2). 9 The monopolar needle electrode is an insulated straight wire electrolytically etched to an uninsulated conical tip that serves as the active surface. The reference electrode is a surface plate or disk electrode placed on the skin, preferably close to the intramuscular electrode to reduce extraneous intervening bioelectric noise. There are a several diameters of monopolar electrodes with the same limitations on electrode stiffness. They have different recording surface areas but there is less information available and basic studies have not been performed with monopolar electrodes. The single fiber EMG needle electrode is based on the same cannula size as the routine concentric electrode but the active electrode comes out the side of the cannula 2.5 mm from the tip and is 25 μm in diameter. The cannula is used as the reference electrode. The macroEMG electrode is a longer single fiber EMG electrode with the distal 15 mm of the cannula uninsulated. The uninsulated 15 mm recording surface is large and can include all or most of a motor unit. Surface electrodes can record the entire motor unit and consists of two electrodes, the active generally placed over the belly of the muscle and the reference near either the insertion or origin of the same muscle. The number of muscle fiber action potentials in the intramuscular MUAP recorded by the commonly used active electrodes depends upon the size of the active recording surface and the proximity to the muscle fibers. Except for the macroEMG electrode, the MUAPs recorded by concentric, monopolar and single fiber electrodes consists of relatively few individual muscle fibers of the motor unit. Concentric needle and single fiber electrodes were used in my studies exclusively. The uptake area from the standard sized concentric needle relative to the size of muscle fibers (average 50-60 microns) and their extracellular action potential voltages has been determined as the 10 Figure 1.6: Decline of peak-peak amplitude as a function of distance of electrode from muscle fiber (11). A multi-electrode needle with an index active electrode used to measure the 0 μm by maximizing the action potential. Subsequent distance meas-urements were made using active electrodes along the axis of the needle with known distances between. Each line represents a different muscle fiber diameter. The difference in amplitudes at 0 μm from the muscle fiber is presumably due to differences in the muscle fiber diameter, but could be due in part to optimization of spatial orientation of the electrode to the muscle fiber. 11 distance where 90% attenuation in signal occurs, and is 350 μm from the electrode face, with negligible contribution from fibers beyond 500 μm (Figure 1.6) (9,10). In the setting of the density of muscle fibers within a motor unit this implies that perhaps only 2-4 fibers' action potentials will make up the largest part of the MUAP and overall only a mere 7-15 fibers will contribute to the waveform (Figure 1.7). As shown earlier, this is a very small portion of the fibers of a motor unit. Further, with random insertions of the electrode different portions of the motor unit will be sampled resulting in different waveforms due to the nonuniform distribution of muscle fibers within the motor unit. In addition, small movements of the electrode will change the contours of the MUAP. Intramuscular electrodes are a blind probe of the motor unit. As mentioned above, the nearest several fibers make up the largest contribution to the shape of the Figure 1.7: Schematic representation of the individual muscle fiber action potentials in relation to proximity to the recording electrode and the MUAP that results. 12 MUAP. These two facts mean that there may be considerable variability of the MUAP from the same motor unit depending upon the chance placement of the electrode (Figure 1.8). This blind approach dictates that many MUAPs must be recorded and statistically analyzed to achieve a representative estimate of muscle fiber architecture. A typical QEMG study requires the measurement of 20 separate MUAPs with no further change in metrics with the collection of additional MUAPs (12,13). Currently there is no technique that allows for the reliable and clinically practical measurement of all muscle fibers of a sufficient number of motor units at one time. Because of attenuation MUAPs recorded from surface electrodes are very small and their size varies with their depth. Further, there is little or no selectivity with overlapping motor units. Intramuscular electrodes minimize the overlap of unit potentials due to Figure 1.8: Schematic representation of various MUAPs of a single motor unit. The proximity of the active electrode surface to the nearest muscle fibers dictates the shape of the resulting MUAP. Therefore, many different MUAP morphologies can be represented from the same motor unit. 13 limited uptake areas and thus pick up potentials from fewer motor units. By positioning the concentric electrode nearer the muscle fibers the potentials are also larger, although the number of fibers contributing to the potentials is a very small percentage of the full motor unit. The single fiber electrode reduces the number of fibers contributing to the potentials measured, but also reduces the clinical usefulness of the information collected. Reducing the electrode size even farther would allow for truly single fiber measurement or even intracellular measurement-although this would be of little use in clinical decision making. We will discuss later, in the section about pathologic changes in the MUAP, why the concentric and monopolar needle provide the most appropriate mix of information for clinical studies, balancing selectivity and comprehensiveness. Changes in the MUAP due to nerve injury: neuropathic motor units Nerve injury leaves muscles with varying degrees of denervation. In cases of incomplete injury of the nerve supplying a muscle the orphaned muscle fibers regain neural control through collateral sprouting of remaining nerve terminal branches (14). This is the process whereby remaining α-motor neurons create new intramuscular sprouts either from a node of Ranvier or near the neuromuscular junction. The degree of retained motor control and strength depends upon the extent of loss of nerve fibers and the time course of loss, monophasic or progressive (slowly or rapidly). The effect of collateral reinnervation is increased likelihood that muscle fibers from the same (reinnervating) motor unit will be adjacent or closer together than occurs normally (15-18). As a consequence, when there are even one or two more muscle fibers 14 within 500 μm of the recording face of the electrode both amplitude and area of the MUAP can increase measurably (Figure 1.9). In extreme cases where very few motor neurons remain, a single α-motor neuron might innervate every muscle fiber within the uptake area of the needle electrode. The pathologic marker of collateral reinnervation is fiber type grouping. The degree of collateral reinnervation of a motor unit is limited; the fiber density increases the radial territory (area) is limited to the original boundaries imposed by muscle fascicles resulting with extreme loss in areas of muscle with many permanently denervated muscle fibers. New axonal sprouts may be smaller in diameter and have less myelination causing slower transmission of the depolarizing signal along terminal branches. This can mean that some muscle fibers of a motor unit will depolarize somewhat later than other fibers Figure 1.9: Glycogen depletion study in the rat after partial crush of the sciatic nerve showing the effects of collateral reinnervation. Note that there are many more immediately adjacent fibers than in the normal muscle (compare to Figure 1.2). Adapted from Kugleberg (16). 15 of the motor unit. In addition, muscle fibers will atrophy after denervation and will conduct muscle fiber action potentials more slowly from the neuromuscular junction to the recording electrode and arrive later than their normal counter parts. When one fiber within the uptake area of the electrode depolarizes at a different time than others it can be seen as an additional peak in the MUAP. The delay seen is inconsistent due to normal variability of alpha motor neuron discharge patterns and the changes in muscle fiber conduction velocities associated with the discharge variability, a phenomenon called velocity recovery function (19). These factors affect the MUAP waveform and result in increases in the number of turns and sometimes increases in the number of phases and a degree of waveform variability from discharge to discharge (see Figure 1.10). Higher than normal number of turns (polyturn; >5 turns) and phases (polyphasic; >four phases) Figure 1.10: Schematic representation of individual muscle fiber action potentials from atrophic/reinnervated unit in relation to proximity to the recording electrode and the polyturn/polyphasic MUAP which results (compare with Figure 1.7). 16 is often called increased MUAP complexity, and suggests that there is some degree of nerve injury in the recent past with active reinnervation. However, polyturn/polyphasic phenomenon can also be seen in myopathic processes, which will be dealt with in the next section. Additionally, newly sprouted nerve terminals and neuromuscular junctions may not have the same robust synaptic connections with the muscle fiber as more established neuromuscular junctions. The acetylcholine receptor is composed of five subunits and with reinnervation the normal gamma subunit is initially replaced by an epsilon subunit that imparts slower and less reliable receptor channel openings (20). When postsynaptic receptors are developing on reinnervated fibers, signals across the neuromuscular junction may be insufficient to cause propagating depolarization of the muscle fiber membrane, which is called transmission blocking. Signals may not block but transmission can be slowed (measured in microseconds and called "jitter" at the neuromuscular junction level). Both of these can lead to discharge to discharge variations in muscle fiber action potentials that contribute to the MUAP. The summed variations in transmission in the MUAP waveform is called "jiggle" and is an important EMG finding that support the diagnosis of ongoing denervation and reinnervation (21). Over time, postsynaptic receptors revert to more secure forms (gamma subunit in place of epsilon subunit) and transmission becomes more stable and jiggle is reduced. When consideration is given to MUAP metrics of amplitude, area, and complexity a reliable estimate can be made of whether there is a high likelihood that an abnormality is due to nerve injury. For example, very high amplitude MUAPs with normal complexity indicates remote nerve injury with sufficient time for stable reinnervation. 17 Normal or very slightly increased amplitude with increased area and complexity and jiggle indicates relatively recent injury with new and active reinnervation. Low amplitude can mean that the electrode is not near enough to fibers from the motor unit, or that there has been denervation without reinnervation and subsequent reduced numbers of muscle fibers. In routine EMG studies these factors are weighed in a subjective manner. However, knowledge of the changes in the various metrics and their implications are based on quantification of metrics and correlations with clinical states. Changes in the MUAP due to muscle fiber injury: myopathic motor units Muscle fiber injury changes the MUAP without necessarily changing the relative cross-sectional position of muscle fibers in the motor unit. Many different diseases of muscle cause changes in the MUAP, including: inflammatory myopathies, metabolic myopathies, inherited dystrophies, and membrane disorders (channelopathies). These conditions result in changes to muscle fibers but the number of nerves reaching the muscle remain normal. The MUAP is affected in the following ways. In cases of inflammatory myopathy the initial insult is often immune mediated, activating the body's immune system to degrade and destroy the muscle fiber at segmental sites along fibers. This leads to smaller diameter and less uniform muscle fibers (22). As propagation of the depolarizing signal is dependent on the radial diameter of the muscle fiber, if different fibers of the same unit have a range of diameters they will propagate signals at different rates. This can, in the extreme, lead to an MUAP where the action potential of each contributing muscle fiber is clearly differentiated from the others. 18 In other words, there is decreased overlap of individual muscle fiber action potentials. This phenomenon causes decreased MUAP amplitude, due to a lack of constructive interference, and increased waveform complexity (23). In some myopathies damage to a muscle fiber is severe enough at a single point along its length that the fiber is no longer electrically contiguous. Loss of propagation of this muscle fiber action potential reduces the amplitude of the MUAP when recorded beyond the point of the damage. This leaves the distal segment of muscle fiber electrically denervated. Fiber atrophy occurs in this segment and it often is destroyed. However, with control of the inflammatory disease process repair is made and the distal segment is reanastamosed with the intact portion of the fiber. This will allow for propagation along the full length of the fiber, though it will be slower through the atrophied portion. This causes decreased amplitude and increased complexity as described earlier. In general, MUAPs from myopathic units can be differentiated from normal units in their relatively low amplitude and markedly increased complexity. While increased complexity occurs in cases of neuropathic damage, the amount of complexity is generally far less than is the case from a myopathy. Also, myopathic units are generally smaller in amplitude and area than normal while neuropathic are almost always larger in both metrics. Though this is generally true, there is a lot of overlap of the distributions of MUAP metrics for normal and diseased states-thus the need for quantitative EMG studies which can more sensitively tease out the differences. Examples of changes due to the different pathologies are shown in Figure 1.11. 19 Figure 1.11: Examples of different characteristic MUAPs. Note that these are photo-graphically isolated MUAP waveforms whereas during routine EMG studies there is a train of MUAPs discharging at different rates with overlap of waveforms leading to the possibility of "apparent" complexity. Row A: stereotypical MUAPs from normal muscles. Row B: MUAPs from myopathic muscles. Note that these are markedly more complex than the normal MUAPs, but that their amplitudes and areas are not comparably increased. Row C: MUAPs from neuropathic muscles. Note that these also show markedly increased complexity as well as much increased amplitudes (clipped in figure) and areas. Single fiber EMG Single fiber EMG is a set of techniques which use a special single fiber electrode to obtain unique information about the muscle and pathology that cannot be obtained otherwise. Due to the 25 μm diameter active recording surface and effective up take radius of less than 300 μm this electrode records from only one to three muscle fibers of a motor unit at a time (24). 20 One of the important studies performed with a single fiber electrode is determination of fiber density. This is done by positioning the active electrode as close as possible to an active muscle fiber-determined by a single muscle fiber action potential of greater and 200 μV amplitude and leading edge rise time of less than 300 ms. When these criteria are met the waveform is examined for evidence of additional muscle fiber action potentials-most commonly a second peak (example shown in Figure 1.12). The number of muscle fibers within the uptake radius of the electrode is recorded for that site, and the process is repeated until a total of 20 sites are sampled. The fiber density is an empiric number and provides information as to the packing density of muscle fibers in motor units. While an increased fiber density is suggestive of neural degeneration with Figure 1.12: Schematic representation of fiber density study using a single fiber EMG electrode. The semicircle represents the 300 μm uptake area. In normal muscle one or two fibers at a single sight may be recorded. In reinnervated muscle many more may be recorded at one site. 21 reinnervation it can also be observed in myopathic disorders due to loss of muscle fibers and greater closeness of remaining fibers. We have shown the characteristic changes which are common to the various categories of pathology measured with intramuscular electrodes. Changes in motor unit amplitude cannot be distinguished with surface or single fiber electrode. Using a surface electrode does not allow for the distinction between either small or large amplitude units because the size of the unit measured from the surface is due to the proximity of the motor unit to the electrode rather than the remodeling of the motor unit architecture as measured intramuscularly. Single fiber electrodes also fail to register changes in motor unit amplitude as they record from, at most, a few fibers and so sensitively that there is rarely constructive overlap of the fiber potentials. Increased MUAP complexity is also difficult to ascertain with surface or single fiber needle electrodes. The complexity in surface potentials is completely obscured by the overlap of many small and low frequency potentials contributing. Single fiber is somewhat more sensitive to complexity, but is limited by the number of fibers in the uptake area. Single fiber EMG can see increased complexity, which is the purpose of a fiber density study, but cannot distinguish complexity differences between myopathic and neuropathic processes as the larger concentric or monopolar needles can. Frequency space of EMG signals Frequency spectra of interference patterns and EMG signals are dependent on a number of factors. For individual action potentials the frequency spectrum is dependent 22 on the proximity of the active face of the electrode to the nearest muscle fibers. When the electrode is close to a muscle fiber the high frequency components will be more pronounced (25,26). Muscle tissue acts as a low-pass filter, filtering out higher frequency components more effectively as distance from the source increases (26,11,27). This is due to the impedance of charged molecule movement through the tissue. In a volume conductor there is no voltage without current, and current is impeded in muscle by various forces. There is natural impedance of ionic fluid, which is far more complex in the biologic environment from increased viscosity due to proteins, cellular and extracellular structural architecture. Cell membranes and connective tissue also impede the movement of ions. The nearness of the electrode to an action potential source (muscle fiber or fibers) can be judged by a rapid rise time of MUAP indicating high frequency components in the waveform. Individual motor unit action potential frequency spectra will depend largely on the overall duration of the potential. Low frequency components will dominate if there is any significant portion of relatively flat baseline included in the analysis (see Figure 1.13). This is also true of interference pattern (sum of many motor unit action potential trains) frequency spectra (see Figure 1.14). When fewer motor units are recruited greater portions of the signal are at the low end of the spectrum. A fully activated muscle will exhibit a shift toward higher frequencies due to the lack of free baseline, though there is still a relatively smooth continuum of frequencies represented. Near the high end of the frequency spectrum the largest influence is the proximity of the electrode to the nearest active muscle fiber. 23 Figure 1.13: Sample MUAP with the corresponding fast Fourier transform spectral analysis. Modified from Pattichis (28). Figure 1.14: Time (upper) and frequency domain (lower) traces of 20% isometric contraction from biceps brachii muscle. Modified from Duchêne (29). 24 EMG Studies EMG is used in a variety of settings, from clinical neurology and physical medicine and rehabilitation to academic research laboratories to visualize and quantify the activity of motor units. The studies included in this work focus on determining factors that influence the physiologic recording and algorithmic processing of the signals to make QEMG more applicable and more easily used in the clinical setting. Clinical EMG Clinical EMG studies are generally performed as part of a suite of electrodiagnostic studies to help arrive at a diagnosis of a disorder affecting the peripheral nervous system. Electrodiagnostic studies are unique clinical tests in that the clinician is actively involved in designing what tests will be performed and performs the tests him/herself, in contrast to, for example, imaging studies or electroencephalographic (EEG) studies, that are performed by technologists and interpreted at a later time by physicians. For electrodiagnostic studies, the clinician interviews the patient to clarify the clinical question and then performs a focused neurologic examination. From this information the appropriate tests are selected. Nerve conduction tests that can assess the function of sensory and motor nerves are usually performed first. The clinical EMG study focuses on the motor system. It is performed by a physician using a needle electrode which is inserted into the patient's muscle, or muscles, of interest. The clinician will listen to the amplified sound that the waveforms produce as well as watch the waveforms on the computer monitor (equivalent to a cathode ray oscilloscope). Of 25 note, listening to the waveforms is a sensitive means of distinguishing differences in amplitude (loudness) and complexity (splitting of sound components). However, this is qualitative with the high chance of over-recognition of rare events (loud and complex MUAPs). Once the clinician has obtained the desired information at a single site, the needle is repositioned a number of times within the muscle to examine a different group of muscle fibers and motor units. As with listening to motor units, there are statistical issues as not all areas of muscle are similarly affected by pathologic processes and a suitable sample must be obtained. Most important for the advantages of QEMG, abnormal MUAPs will be admixed with normal MUAPs, and a suitable sample size is necessary to avoid false positive and negative interpretations. Many EMG studies require examining several muscles, repeating the process described. Clinical EMG studies are performed by clinicians who have been trained to recognize normal signals and differentiate them from abnormal. The patterns of abnormality discussed above are often very obvious to the clinician, who can then use this data to arrive at some clinical conclusion. Of note, the well trained ear can distinguish very subtle differences in MUAP waveform components, including component frequencies, changes in complexity (number of phases and turns), differences in timing of components (jiggle), and overall discharge rates. If, however, the EMG exhibits only mild abnormality there can be ambiguity for even an experienced clinician. Thus there is a need to increase the sensitivity of the studies performed in the clinic. 26 Quantitative EMG Basic mechanics of QEMG analysis QEMG is a term that broadly describes any of a number of techniques that attempt to quantify some aspect of the EMG signal. This body of work deals with quantitative multi-MUAP analysis. In the early days of EMG studies many attempts were made to quantitatively measure the waveforms, but were extremely tedious and time consuming. The earliest quantitative studies were performed by photographing waveforms from an oscilloscope and were measured by hand using calipers. This provided the MUAP data upon which qualitative EMG studies are based, but are not practical for anything but research purposes. Modern computing has made the QEMG study possible in a time domain that makes the analysis practical and also allows for determination of derived metrics such as waveform area and area-amplitude ratio, assessment of jiggle and fiber density. QEMG focuses on the identification of individual MUAPs from a weak interference signal that includes the activation of several motor units. These techniques are called multi-MUP (for multiple motor unit potential analysis) or decomposition EMG techniques, based in part on different algorithmic approaches. The signal containing many MUAPs is most often analyzed using template matching (30-34). The different MUAPs are characterized and compared to determine if each recurring discharge is likely to belong to a single motor unit or several. Template matching is performed by assigning an identifier to each extracted waveform isolated from baseline. After these waveforms are identified they are compared to each other to attempt to match them based on similar metrics and matching 27 morphologies. If a waveform is sufficiently similar, it is assumed to originate from the same motor unit (see Figure 1.15). This determination is strengthened by analyzing the discharge pattern of the motor unit. Normal physiologic activation of motor units falls within a range of frequencies from approximately 8-20 Hz during a moderate contraction (35,36). During maximal contraction, which produces an interference pattern that is too complex for QEMG studies at this point, the motor neuron can fire with bursts of 60-140 Hz (37). This relatively regular motor unit firing pattern, during moderate contractions, allows for predicting a time window during which the same motor unit would be expected to discharge. Some algorithms exclude all waveforms that occur sooner than expected from Figure 1.15: Stylized example of decomposition using template matching (32). In this drawing, recurring MUAPs are assigned an identifier while non-recurring waveforms are not-they are assumed to be several MUAPs overlapping. 28 the previous discharge due to the fact that it is extremely unlikely to be physiologically stimulated at such a close interval. Once grouped into sets of similar waveform shapes, the many discharges of the various MUAPs are respectively averaged to remove random electric and physiologic noise. Averaged MUAPs are then automatically measured for each of the metrics of interest. Waveform marking is performed by algorithms that first indentify the onset and termination times of the MUAP. Once these two points are set, amplitude, area (and derived metrics such as area/amplitude ratio), and number of turns and phases is fairly straightforward. There are different approaches to identifying the onset and termination (Figure 1.16), and small variations in threshold settings can markedly affect the results (12). Generally accepted practice is to collect a number of trains of EMG signal in order to obtain a minimum of 20 averaged MUAPs for statistical rigor (38). Advantages of QEMG Quantitative EMG can become a more consistent and sensitive tool than the qualitative clinical EMG study. In most pathologic conditions the distinguishing motor units are in the minority (Figure 1.17). Subtle differences can be discovered in the QEMG study that may be overlooked in a qualitative study. Conversely, given the bias of the clinical exam, a qualitative study may overstate or understate any perceived abnormalities inappropriately where a QEMG study will not be biased. 29 Figure 1.16: Various algorithmic methods for marking onset of MUAPs (12). Each example shows a trigger level which initiates the search for the initial excursion from baseline where the initial duration marker will be set. Methods A, B and C each work backward from the site of waveform onset to ensure that no earlier waveform component exists. Methods D and E work forward from some point a set distance far in front of the trigger to catch the earliest waveform component. Small changes in the values used in any of these algorithms result in large differences in the resulting placement of duration markers. 30 Figure 1.17: Distribution of MUAP metrics for normal muscle and disease states (39). The figure shows clearly that there is significant overlap of metrics measured and it is the outlying values that help to distinguish the pathologic from normal muscles. Practical issues Currently, very few clinicians use QEMG as part of their daily practice. The reasons range from ignorance of the techniques or lack of software to perform the analysis to dissatisfaction with the extra time the quantitative studies require. Though QEMG does legitimately take longer than a routine clinical examination, with refinements in the algorithms the time required is decreasing and will soon be only minimally longer. Patient discomfort (and hence tolerance to the study), though not the primary consideration in any medical procedure, is important. Clinical observation shows that there is a full spectrum of pain and anxiety associated with EMG studies. An advantage of QEMG is that an informative study can be performed by investigating a few key muscles. 31 EMG machines, like all medical devices, require approval through the FDA. This means that each new component of the system, either hardware or software, requires an approval process which is lengthy and expensive. Therefore, if a component of the system could be improved there would have to be some financial incentive for the manufacturer to go through the process of development, testing, and approval before it would be available to the clinician. This argues for the improvement of QEMG techniques using tools readily available within the existing EMG systems so that such improvements can readily be employed broadly. 32 2 RATIONALE FOR THE WORK AND TECHNIQUES EMPLOYED Bringing QEMG to the clinical laboratory is the ultimate goal for most investigators involved in the development of algorithms and new techniques. However, those who develop the algorithms are generally computer scientists, programmers, and basic scientists. Thus, there are many issues in the clinical realm not appreciated or identified and left unresolved. Without direct clinical experience the practical importance of certain aspects of QEMG may be inappropriately inflated or minimized. We can bridge this gap because of our experience in the clinic and background in physiology, engineering and biomedical research. We have explored some clinically practical problems from an engineering perspective. The following studies address practical and theoretical questions that arose from clinical use of QEMG techniques and different algorithms. 1) MUAP detection algorithms are empirically based and some algorithms seemed more efficient than others: however, we found no such comparison in the literature, and thus we performed a comparison study to determine the performance of each. 2) Measurement of fiber density requires a special single fiber electrode and we investigated whether the same information could be obtained using simpler and more accessible QEMG techniques. 3) Basic physiology of signal propagation led us to investigate possible diagnostically significant differences in MUAP morphology when recorded near and far from the site of 33 the signal initiation (neuromuscular junctions). 4) There are two sizes of concentric needle electrodes available and the smaller size has been used for studies of neuromuscular jitter, and thus we studied whether there were significant differences with respect to MUAP metrics between the standard and smaller sizes to determine if the spectrum of QEMG could be expanded with the use of one electrode. 5) Recording conditions change over data acquisition time due to physiologic and adventitious movements of the patient and operator and we sought to determine the effects on signal stability on the time trying to keep the needle electrode in one position within the muscle. In the following sections we also lay out the basic techniques used for each study. A more detailed description of the methods is found in the individual publications, though we also include additional information not found in the articles due to space restrictions. Algorithm Comparison A number of different computer algorithms are available on commercial and research EMG machines intended to perform QEMG studies and we undertook a study to evaluate these algorithms' efficiency and accuracy. The algorithms are proprietary and specific to the EMG machine. Heretofore there has been no direct comparison to guide the selection of an EMG machine. Since QEMG provides statistical data it offers the possibility of more sensitive analysis and comparison of data from different clinics. This can only be realized if the performance of the data collection devices is comparable. In this study we evaluated three different algorithms, two commercial and one in 34 development (that was soon after commercialized) against simulated signals and biologically obtained EMG signals (40). We tested the algorithms, one against another, using biological signals because this is the intended use. We brought all three EMG machines together and split the signal from the electrode three ways to allow each machine to capture the exact same data. We could not obtain input impedance values from the manufacturers of the various machines, but did determine that there was little or no decrement of the signal when multiple machines were connected. We collected data from normal (healthy) muscles, and muscles from a patient with amyotrophic lateral sclerosis (a neurogenic condition). We felt it important to include a range of MUAP morphologies to more fully test the algorithms. Each of the machines was also fed simulated data (trains of simulated MUAPs) so that we could compare their performance against a set of known values. We sought to create as realistic a test as possible. First we created simulated MUAPs by recording from a variable power supply through an analog-to-digital converter. The waveforms were produced by manually adjusting the output of the power supply over time. The resultant waveforms were imported into MATLAB (MathWorks, Natick, Massachusetts) and adjusted to even more closely mimic biologic MUAPs. Twenty-eight different waveforms were selected that were similar to naturally occurring MUAPs, and each was analyzed to determine standard metrics. They were mixed together in various combinations. Simulation of a normal interference pattern required that we mimic discharge frequencies of each individual motor unit independent one from another. From each 35 MUAP a train of discharges was created. The mean frequency of discharge is known to be 8-15 Hz for healthy muscle (41). There is a nearly normal distribution of actual frequency variation around these central frequencies. We chose a single central discharge frequency for each of the simulated MUAPs and used a normal random number generator in MATLAB to create variability of the interdischarge interval. Each of the resulting distributions of discharge intervals was then compared with biologic data. The distributions were statistically similar to biologic data. Each train, consisting of the firing of a single MUAP repeatedly over 30 seconds, was saved and combined to make interference patterns. The creation of interference patterns, the signal resulting from several MUAPs discharging near the electrode, was accomplished by simply adding several of the previously created trains together. These voltages are known to add algebraically. This resulted in a realistic interference pattern where many discharges of each of the MUAPs had free baseline on either side and occasional overlapping of one or more MUAPs resulting in both constructive and destructive interference, creating a unique waveform that did not recur repeatedly in the interference pattern. The interference patterns were played back as if a routine clinical examination were performed and we determined that they were virtually indistinguishable from biologic data. Figure 2.1 shows a simplified illustration of the process used with the simulated data. The obvious advantage of the simulated interference patterns in the algorithm comparison is that the precise number of MUAPs and their metrics are known. 36 Figure 2.1: Illustration of how single MUAPs are incorporated into trains of single MUAPs then full interference patterns which are then presented to the different EMG systems. 37 Comparison of Three Algorithms for Multi- Motor Unit Detection and Waveform Marking Reprinted with permission from Muscle & Nerve 33: 538-545 38 , • ; ~ f i , ,,,•• ~ 1; ,! ; I ,• ! ,, ; ~ ~ ; <", J ,i ~ ·- ~ ~, ; ~ ':;. , "• "• •, ,1 ;~' ">, , . · , ", ; ~ i • , • " ~ , ; 1 ,: . =~ ~ ;, ~ - ; .' ,I ". " '.. ;',~.i,~-.o , : T , ;, ~ ! ol , ~ , : , t", '~~ 'c~ -_~' l. 'oc~~-" ~"· ·,~ ,~.. "; · .-." ~ ' r·~:" 2 " , ' '', ' "~~ '";-~~/,'")\~-',. -; "~" ~; "~-~2. ":~ ",t. ,~,:,"<~~ ."-\,h..,' "-~'<1"-"~.; ,~''~~~~;:r- --!-1~'. ,,:; 1c,,~I;-""~., ~rv ,.:; ~,~ ~'{- ""t" . "'- -~,," r~ "" ~""_o r~- '~~,'~ "F ~,Ji' " . '" l ." ,. ; ~ '; '" ~ _ , , _, ;'';~ '".,C~ ~1,-" '}•-"•'l" ,'-_ ,~'.~,1""",.·0"";", . "•_. '_l~,,~_)~!r.-,~~_-_,~ .,y",{" ,~~~~-,.~'," t,,_ ~_ I. ', ;. ,, ~ . ~~" - : ~ - ~, ". , .",- )"""> '~- fl!' "c :'i', 'e r~" . . . , ' < ! ~ ;- ~ 'I ", r ,~ '-- ;0 ~ ~ ~ ,- I ~ " . , ~ " ~ '. ~ t r ! r ~ ~ , n~ ; -, ,~ I~ , " , ~l ,,' !~. ~. ;__ dc ; ) ~ ;~ i>~tl( n:' ~q' "f ~H,.; ~', " ~"~ ~~' " .. ; ~ . ~ ~ . ~ ~ •. J~" ~ . , t c ,,~ o, . ~ ~~, ~ " - ' .. . l ~ · " t " 7H; ,'< ie" ,; . •" ~C' 'l"" '\- "o ::.~! ~f~' '"~ ' ' r ~ " . j . • , _ ! ~- ~~-' f;. ;- ·. ,". .H ;~ .o ~E'~l " ~'~ ~ ~ '! ' ~ <, ~" ~ c , ' [ ' . ~ '~." -: "~p. ~1 ",,~""~," "~."<"". , "<"~j,- -"' :- '~~ ,.' .,,~- 1 •p i p• \' ild '~ :Tt • jj' ! ,j ; "~I , j 1 , " ,• • ,,• I, f I , , "" . H, , ! ' I i •• •• C,,I ~.0n0 •0•" , .0' <r-! !:I:'8 0" '00 ;g~ .0 • ~iElII i-l"' 00' ., "0 ~~ .~ 0= .. • ~ ~ • ,',i 1 39 • ,,' • -- "- 0 ... , 0 • •• , ,,, ,, il--'- , ,,, -"'."_ ,,"-'" , , '" "" '" "" '" "" - ~ ""., ~ """.~ ,"', ,~. ,.u ,·""· ' ''" , .j" CO" - ,~ ~" ",, _, ; , " ,_, . "w , ~ ,,... - " " ~ '-' , ~ , '"," ' , . " ~~-. " ._., ~",,..· · - '." V ~' ,.." ~. " ,' ,. ', ,o~_"~ ' .. .... n." ; "< ' ~ '" " ~"""'" .- , 0 0 " , , ~ , Ii , ' 0 -- , U'" ." , ""M, , ,-, , --,- , -- - ,. ,. " -, r,', '0" -" ,h' .. ,,1 ••. , ;"''',' .'' " ,""c " ,. N"C_' ,, ' ~ ,, ' - ,, ;, .~ ~ - ro N'"' ', ,,, ,.. " "'T ,"_" , _'k " . < ., .. ,, . " "", ,, , ,, -,,. Vi.·" r>, ~ ', ." = '" 40 i ,~ • i , ~ • ~ I t "• '<• ",' ) , · ,~ , • •. ~ i , ! " <; b "' ' • ,.'" y,.~ ,, ~I ~ · ~,o '~~" ",~ " 1 '" ""' ~7" '> ~.~~~l­;';~ " ~ I ' , ; §1, ~ h ,rc " ~ - , --,' ~ ; ; " '',''-.''ti' ,,' ;- •"." • :I " hi;,~ ~."~. .- ~ i ' _. i" " '" ' . ., • -f' • • • • • t -) :.- 1", ,, .. n, • : - ~ , o ~ !;" '-- 'n~ .,l.r ,; " ~ ~, ~ < 1, q ~ ~ "' - • ••• . • '. . ~ , - ~~" ' F " "~ _' , :'~;- ~ ~ ~ ~~i , "~ '; '" .- "H.c ­', ,,; .. , ~- .. , ~~~ ; p~~ ~ !l-L , . . '" . " , I~ 10 Ii ., ~',= ~ ~ • ".- • ~' . .,. • • • ~' - " Z ,... ,- ,p ' "~ ", ,-. ,.' ,o 1 "c • -~ ,~. '.i "~ .>.. [ -, . " ,~ . , ,J ~ " ~ " Y ~ • . _ . ~ "q '~V<~' .h,," jt" "'~ l' ._" J ~,-F• ~!)vo ~,'~~.~_' !~1rH 'F~-~.n- "l ~t~\ ,i i ~: ~ : : ;l~~~~:;l ;r¥"" ~~~r?i ; i ~ ;~~t. ~~~J~ ~:~ ~ ; ~~ ~l ?p." ~;!t3:~; h~~~ ~t : :' iI' ;;.,'}< I, ...,~ '' "< -'d~f',1'~ ~- "-'-~'~ ~.-~ ,~ , ;~_ .i, "" "'-~~'~p "'";'C---<~l '' .1 ' ;~ i l ~, :':j'~ t o; :t' Fo~: ~ ~ n' ;; :-1', j l. '1 ," til' '"i." ~,- ~ ,~ -. , . - , ~ " .. , '.' "c?~ ~ l,~ .~. ,"" ' , _" • • ,," ~ , ',0 l"~ .• " • . ; • . - ~' .,-.~ , - ~ "- . .. ~ -"- .. .. - . ~ . ,' ;,..;" ... <'c ' ,, ;~,q~ , ' · ,,.,~,,. ~, l' ~' , ~ ". '1 ",. ~, , <~ ~- . I, ', - •. ' . > ' ~ " '~' l - '~. j,[1_~ n''~'-$1: ~ r:,..n, ", ',~f)_:,-;, !-~q. •; ' 1q;\' l<itj,",, ,~ ~ '~~<:cl'" _~", ~~ " '[<~':~' ~ ' _" , ' " , ~ l ' ~ ", , ~" ,",1' ~' .' ;' l" ~ · ; ~ ~ 'i~'H,t-,- <i,'HI ; , ~q f' ; ·" _ •• . E --l"'~ ' •. ,-- ' .",. ~ ,T ~c,~ ~ ~~ i.; - ~'~~. , _11 , Z \:;~,l~; ~' ,; ,'~~."-'~--,',~-~, ', - , .,~. t >t .. ""'--' ' ~--" " . ~ . ".~, ". .~ . ~ .; "O _YP,d. ,. - ~~, ; . ; ' to ."• -" ."" "c ' __. " • '" - ,­'-." , ~ , ~ . , ,, , ,-,,,~ ,-- 1",'" I ': ~, ~'"i-0",'~, i : < ~ o :, . ~ ~' "i~=~ ,:~"~,, ,~ #, .~~ , ., "'···'-.f"'fl~'·e" .,. ,. ~; " .. ,. "',,-"~[" ;-.- ; ,~ .ii" ;: ~ < . ~ " ., u . ~, , 41 ~ ~ o ! ,! ! ~ ! I , , ~' l " ,'_ .' ; ) ~ .. ,, - ,. ~, .~ ,' , ~" 1 .~ "~j,'.,;.' , I ' L . ~: t ~ ~ ~ . • H , . ." ,.<,, ;•. •. ,~ '.~. .".'. ~ " < ~.,. . '~: "j · .·· ,!_ ~. ," '·"1"·"·'"· I ;.,.-. ~ · · t;-, .~. ,o .·""'~~t, , ;~, .. ' < ~~:. , >~.' ~ ~i""~" t< . '.; lli~~ "-i , . ' n " l t " -. - ~ ' ,. ,. " j~t; ~ ~ ~.,o - . , , • . ,. ~~~ >;, ~ ,."""1,c ..~ ,,,~) ' e, ,~~;. y.;· _ J ;~'"·., ~~i ,~ , ~ ." :tH o f ~ t ~ " ; -' lc ' r, l ,c· ; ,. , ~. r,.­! i : , H , j t: " ? n • . _. ,.'" " o ~ 'llL" _ ~':; ~ , '". ' i : 1 L~ ; ! .' ~, A ' ;' ,~ . . _", ... -. - ~ . ;, '; ~ ~Jl.l.ti~~ - ... . ~ 'I. "'C ··:·" I· ," .!. •'' ,- ,'.',~-· ~ s: • , o· ~ t' , , ,l- f I ,, " , . jr 0 , "F ; >" > " " ­, • , ' C , .' l' ",. , , ~ ~ , . , ~ , ,.. . ), ., ~ .. , •~, , ~ f ~ . r ' e e. ,H , r, , ,~ "" ~" .. ~, ~ , ~ ~>~ . ~, , " I ! { ",0~' "'. ' ~"~'~r~~ ; ., ~ -, . "" . - " .' "'- ~ ' t~_~ t ' "' ,,, , ~, ~ ( '" , > ' , '- ' " ~ d · l ~.- .. . ~~ ~. - "~~ . ~ '; ~i~' : ' , ~ t ~- : ~t ; < ' ~ ~ t:-: t i ~ ~ J .. ... ' : ~' < . o.,. ,. ~ " !~ ~ j ~ ' j ; '-; L~r; ~ , H . ; ' J , . ~ .. ' . ~ ~ , t~'I~'~ .;: ~ "~ '- "'~" ' ~ ~-',1 't" " ", r • , < , iT' ~~. 7 ~" , ' - "'" Hf' H f.i • ><• ~ .. ,,~ ,:- :q .. , ,• , i ;f~ " " ,~, ' 0. " ~ " "'-' a i ": H i ,.. , .o,,. .~.l .', " '_. ,, ., -f d~ ; , . - .H. " ., - . c. , '. -. ' 1 " , : :~~. ~ qi,)~, ;,.j " ' ; , "'- 1,., ,"> il , Ii l .r. H .Jl~' ~- ,~ ",.. .. , . , .,"- ..- H E"g ". ;'. I- ">o, . ,-"'"" " "'''"'','', '': ;1'" .:~ 'is if;: ! ....... - "'',!1"4~ .•-. ~ .'~ f. ' ..; i •~ f IIi <bl • : . r· ' ,'i ,~ '@........... ~.' .....~ ...'. "'! " , " 0" ~~1~ ~ :s ~f ~I ... .i '!, -. .~' -- ,. "-, 1 I:';' : .~.... .. :f..'..:. :~' . "- ~!~~ ~ : : . ' i ,. 1.-,! -- .~ ". ~ -",, I "< ~,""!'lg .;,--; ; "~ '-I.. ", ," ! !3J'<" 1"_' I ' • .• ri'"'' ._ [ i"~" ,i .~': !~I! .'-~. ". ~. !'"l lJS,?"-o " 9 r III >}I , .·,I_ ' !i~; .-~. , ~ 'r":"l":" ,. .. " ~ ." . 0 - __. ! ~~f - -~ :f ' :11 ~: .I. ; '"" -. ,' ,; , _,'1., '!,~ "M ~ " " . .. ,,, .. <~.".I . .-. ~' "' .~. ~" " '~.~ "" , 't ,~"! '..: ~ of' 'I r L~ · -'.- .' -. ,' ~ 42 '" • ,• j , ,,, , , , ,,, • ,• • ! ,• ,• , ,• , ,• , • ," "' , r • , , , , • , .l' iIJ ~ ,; ". ',;:" ; '; " ., ;,1. ,H. ", ., ," "~ " d i~ 43 I n " ~~<C • . ,-"" f " . . ... ~ 0''' .'-'' ""'"" """ • =r~ .'" . ,., , .• '" , ,,_ , ~",,,~ c J'" Ml ,". " " ',h-,', ', ' ~'_ ~ M ~ ". , ,_, ~'. " ~_''' .'''~ 'W" " O '" "" , ' " "'" , ~, -,H .-, ,,; , - m l . ~· IT'< , "._,,, 'HI·" ,',,-, I" -~<."". d, ' ~~ "" " . ~ ' . ", _,-'-_ ,,,- ~ ",J ",,- ,.~ ~ " " . ,,~,. , ., ... ~" h,' , ,~" , ., cI " . ", IT """'." '" ,,,~, ' h __ - ~" ..." ;-1,.". G _ .. " ..... ,.,,_ ~ ," "'. Ln ""","'0 , ~ ~ ,<;, -n W¥ ,"' 0 " .'''' ",,,,,. ", ",." ,, ~ \iI "'" . ~, '-' -n"", ' ~'-"'"" .. , ''' ~, ., .n .·,. Co.. ,J ~' _. , ~,,_ "' "' ... " 'I C" ,_,·d · ' " ,I -I"M,,.,,·,rr,,,, ,"" -,, ,'," 'M'­..... " " " ,,' II ', 'd , ~" , - "' .. b, .mll'>,'·' " ,'A' Lu.o. _ ~ ~.dm C'. " .• L~ A "'-.0 _ • . """,_ ~ ~ " ,. " "l~~" \11 ' ,I>. \( ~ -'" , ",_., ,., • , ,, ~ ,mIl'> -,., ." .' ,., .. "u~,. ~ _ ,', _~lA L' ~, Lkx ~ . m," ' , ~, OA ' " ,. , ~,_ . ~ ~~",o ,,' , ' Co , ~ ~ '",_,." , 'P"' , ~, "",-"., -" ~, "" • ~~. " • HI·'", "," ,,;,., '0 " """"" " , ' 0_ ..... " '. ""-lA' l-- ,L.· , ... '._,~ Jc... ""-" . ,., Tr, '" ,,~... • n "'" , -.. , ,. "'~, . re. ",,' "~ " ' ~ " 'O , ,,,~ ~ .. ~ ,,,,,,. ,~ ., \(,-", I" II ' C . " " , A~." d ~ ,. , " ,,-,., " .... " .,',', _~LAl' " ~ •. . , 0_ ~"_~LA! ' " ",~. ",,-, .. ~" . ''" .. ""," ,,-, ,,- , " . , ~ , '0 ', ., -, ," h " , ',- . Y" ~ '.- ~ " ,j -, ,, ,",,;.,,- -, ." .. , •. '[' .• ~L.· , , ...... .. " ," '"U.!' ,_"-' ," ' _'-~ . ,~ .~ ,,_,. r ",. r~" em ,. "''' r'-'" '''' W~ ". . , , ~ ... -. . .... " f" .,· ,,'_ 'i ' .'" ' -'"" I ~ wm " " ~,"~ ' .'L, ,,,,," , _'-"~ . " ~_,". "-" , _""", , '--" . , -, ,-, ,, "'" n~" \11 ' \l' '"' ,. c,_ , ,,- "' " ,~. ,;.., .", ,. " , .-+-, d " . ,. • .•• C " . ~, ,,, ' , " " ~, ~",., II ', ,-n,-,",", ,, -I~ ~~'u " -'-"'-'-.- ,'_ ' !lAL', , ..... ' _, ul ~"" ,. ~ ",.", """., '-' , "~" ,-, -." ,.-r.-, , .,. " Q. ' " " .. , "" oj, , .'t,,," ,., , I. ~ ." I d _ , <--,.- i..-~ , ~ ,'''~ . 1.'. ~."-"'-~ " .uAA_, "'._ co ~ ",,'.' ,,-, h " ["-.,,. 'n , .... "" " , ". 1. ~ "' ,, ~ ,',', ,. ,"'- -'" " .~ ~ . Ii • .,~ , , , "' 0_ ,,, .• . , ~ d l .. ~ , 0_ ~_. ,'_ ,L.· ~ I ~.· , . n_ " • . , ,"' ,"" , "''' r, r" .. ,,-, 1f'<. T.,,, ,,, ~ ,h" ,., ..,-~ ' • .•• , .,. \H. '" " -~ , " ,...-•. " \i l. ,, ' • • ~_ - " .... . , 1 "", ,. Ire," 'M" ' '' -'~ ', ".~, ,-,,, ~., o _~U1' , l~ " .~,,, \(~" ~ 'H , .. ",," " ~ '"" '''~ • , .• ;, -'" ." -I , - ~ . ,~ ,,~ , , ~w"'" d _ ,~ ... ." .. , "" .. , . ..o ~ _ ~,~ ul . _ ul " . . ........ ·1 d~· . ' JJ~. 1.' ,"M " " " " ~., ",, C" .• 0 ~ - ,.-, , .,,, M _ " ¥~ d ¥ ,'~ M,·,' ""'" _ ~ ~'''' ', ~,, ' . "",-,,,., 0'" ',,,_ , ,,~ ~ ... J..<b· ~ ,· , l~ 1.' ~ ~ c"" co - ~ d, . " •. ,,~ ,·.' .- .,-'" " d~, ' ~ I ~"~'" , I'" Y_ .~ I ",. " ,~ " ",," .... ~' '' . ", ,,-L--A~ 0''' ...... ' "'_,._ ,_,. ~ ,., ~ ~"" " "~ ''' ~ '' '. , ", .... >" ~.'C , ".". r ... "", ''< ' , .• ;, -'" ,,,, '-' ~ . , , ~ , ~ . '" " .,, ­'" -,,, ,,' " , .", -, ,." , .. , " ." "" ,· , ~· I "",'.­'~" , ~ ·W~""'~'''''' .• _"'-'-, . ""-'~' '' . '' L, . .. j] '~ , II' co" .. '" ' .... ,'" " I ' " ', <,' ' , '. , d.,,", ~ , ., . ,~ "~' " ._.1 ,." ,',h-, ,.,. '''' '"' ,_ ,,," .. ' ._ •• ' , • ••• , , "· .,, ~ .. ·o ""k.~, •. ~~ " . ~ ,,, ~. ,~ (> , " " " , ,, ·. e~ "'" " - ~~ , ~, ,. " _ " '''_" ... .. , ~ "_ " C, 1.' ~~L~ , .... ""-'~. C ~" N' 0 k , ,,,, , ., ,h ' ""..>; - " 'f""N' " ,." . -, " ~ om , .. " ,-,',_ ,'~. " , _ ,,~ ,' .r ," I, , _ • . > .. , ~,,' ~,""' ," ""''' , ", .,,""' "'~ ,,, ,~.~ ="", ., ,~ , ,"u,, ', ~"' '' '''' ,"" .. " ~" .. . ,. ~ ." ","" ,'=.'''''.-h', ,"-.-."--,,.-, .,," "'''' , ... , "" , .. ,.- " ..,'",",,' c,," • "" " " .. """ ,~ .'", " ",' '" ,., ''', ,." . .. , .. " .,,", C, " '" •"• ""'" ""., 44 ~,. " ,,,. :t~ ]" ")')'" LL . '\ t,· ... , ,·,· ... 1""',-· .. 1 .. ~ .. , , • •. , . "v ~.,."" , • ,'0·"" 1,,'0-· , , -"" , • .• • ~ ~, " . • :~ , , " )1\4 ~. , " . ~ """ _ , ... ~ ~ ~, ~l '-,,--,n "''-''·Y ',11 y , ~ """( ,- ~ ,_, ''''0' ,.-" ' U " . ' "<, "";:, , _. ~ .. " A ' '-' ~ 'W .' '', ' OT T ''T'' ' ,. ,,~, ., ' .~" ." " " '.~~ .~ " .• , 'r ., v .• ~ Ju· "l' ,0- , ,_ ' ''~ .,~,- r<Y,' m~ '''r "'p r" ~_._, _",. ">.:>0 ... • "". ,\ . • .. , , • li9>·o . .... p'o·o- . • "" O~'V , _" " . ' "' '_ ~ , .-.... "" ~ "w .... , ,,~ "'-", ,. ~l '" '< ~ ~ ."'''~ "On .,- p.Y~ ' "" " .­,, ~"'W " ' " ,~, ... _",~"." , .J. ,." ,_, ' ''1' -'H"'" ·'T .',"", ~- , - , ~ . - o,, ~ ~,,- ' " <. '" ~ u 'I ~' '._"' "'V, '. " ,~ , " ,. ., .• ,.' .''''p .• , 'or ,.",<.---" ''IT , .'" , ' "",," .-.c ' 4 . '~ __ '.' ""'- 0" ' _ ~ ,_~.w_ ,,,, ',,, ,_'N' ._ .••. _ •. _ -"'-.. _ ._ .• _0'. 0" __ .' "»' ," "., ., "'., .".,, ''''' ~, ' ;n ., .JO .·m· " .,., ;n " '1 en ~ , . , .. m . , " "'0·' '''' . , n .~, ....,,,. , "' ,n " ~"" ," .. " ....,." ~, "., --.c. . ,-,,,,. , "., .U ~"" " " "' " , . , " 0."" ."-."_, . ',~ ) " --' .... .. ".-. .-.... , . , '" ""j.' .' ,-, . , "., '" " '.",,,. . "' " 'r , ... , . , " . , "., •• "' ,qcc·,rto. "p-. ~' " " .V ~ " - .. ,,' ,. .. " "-'.' ~. : . · ... 0 ,-"'," '" ~ ,, ' , .. ". ' '. ' . , ~. ~. ', . , . """ " ~ , ... , 45 ._ ~ Iv ,.~ .......... , ,. ~ ..... ~ ._ .. .,. . ,, . ........... ~. .. "'~.,. .,~ ." ,. " .... ","u' ·.~ ... ~ _w~, ,',.. ."" "'~g. ~ ~~-. --- ..... ... - .. ~ . _. '_J'~ '''~''''' _A''~ ,....~.. ~"" _ '.MW··''''''''''' ' ...... _., ... ,', " .-~ " ... -, " ... -- ; , ,, _ .~ •• ,,' L. •. ,,"' ....... ,', " , " ,'." '_ N' • "-,._"" " ,_,_ , _,,.~ ,~ ~ __ • -. -~ .,."<- ., "" " , .... ...., <- ~ ,, ~~­"" " --'- - " ,... ', , ..... , "--" ~ .. ~. ~ .. '- .. .. ~ .,...---.. ~-- ..... '"- ..... _. _ _ __ .._.- . ....-. '. o<~ -~-.. ..,. , •_ .- .., '<_ A_'J' _ ......, ,_. _ .~.,-_ .......... _... .. ~.- - 00 __ ~~, ~." R"''', '' • '"'U_ .......... _ ._ _ ._'~ ~~~ .• -.~- .. - ... . _,.,..,., .".."..'.'.'..-..- -'..-'". .-..". -'~ " ...... . ','e _' ... _ •. __"' ~.,. . .".'." . "~""'" . ,"".""" .".'" . . ~,.,.'. -..; ", ,' . " .. ...... h .'..,. _ , ~·O ...... - - " ". ,,~.U .• • ,,,,._ .... , .," .. ~ ........, ~ -"" .... , , .. '""'-~' -'''''' '" ,"n0"r. ..". '. _,., "_ ._.. ".., , ." ~" ..•., " ..~... ,, .. _, • .•• •" , .-.....R.. . ' ... _'~ ",.~ ...... ,"'.,~.~ ' ......... ~ ..... _ ,. _~, ........ C"_' __ ._ ---... ........... ___ ... , ..... ' d '''-'"'''' --..: ,, ~,.... ... 46 High-pass Filtering Electrical signals from muscle fiber action potentials are attenuated by the surrounding tissues. The rapid rise times of muscle fiber potentials close to the electrode can be brought out if low rise times from more distant potentials are filtered out. Previous studies have suggested that high-pass filtering at various cutoff frequencies (1,600-3,200 Hz) can reveal increased complexity within the MUAP that may be clinically relevant (42). Such signal filtering could be useful, as most EMG machines are already equipped with high-pass filters that can approach these numbers. If we could determine some utility of performing such filtering it could then be immediately implemented in the EMG laboratory. We undertook a study to determine if the MUAP metrics of filtered signals offered any insight into the physiology, anatomy, or pathology of the muscle (43). MUAPs were recorded from healthy and diseased muscle initially with default filter settings (band-pass from 10-10,000 Hz). These were initially analyzed by the system algorithm (DQEMG), and metric marker adjustments were made manually as necessary. The signals were then exported to MATLAB to be filtered digitally. We initially filtered with a 1st order digital Butterworth filter. We did this because it would be the simplest implementation, and all machines could match or exceed the performance of this filter design. Subsequently, we investigated the effects of filtering with more aggressive and sophisticated digital filters, but the results were not significantly changed. After filtering, the data were reintroduced to the DQEMG program. This was done two ways to determine which method was best. First, we introduced the data as if they had just been recorded. Second, we used the information determined from unfiltered data 47 about the firing times of MUAPs, so that the algorithm need not try to decompose the signal, just average the MUAPs. The two different methods yielded nearly identical results, and we introduced all the data presented as de novo. In the initial version of this study we attempted to correlate high-pass filtered turn count with fiber density measurements. The hope was that we could use a conventional concentric needle electrode, which is much less expensive and is disposable, to glean a portion of the same information available with a single fiber needle determination of fiber density. We collected fiber densities for each of the muscles studied in the high-pass filter study. Our initial data looked extremely promising, but as we collected data from more muscles the correlation became very poor. Ultimately we concluded that this portion of the study was not feasible. We measured the number of turns and phases using several different threshold values of the change in voltage direction: 50 μV which many laboratories use, the more sensitive 25 μV value that some laboratoies use, and all visible turns that we as experienced electromyographers felt were not noise (approx 2-6 μV depending on the noise in the signal). These different groups of data were evaluated against fiber density and it was determined that the best fit (signal to noise ratio) to the initial data was with the 25 μV threshold. We also performed simulation studies to evaluate the effects of filtering. Two models were used (44-46). The Stålberg and Karlsson model is commercially available and the Hamiltron-Wright and Stashuk model is available from Stashuk (personal gift). Correlations between filtered turn and phase counts and fiber density were relatively good for the simulated data. The models allowed for simulations of the clinical single 48 fiber study. The Stålberg model allows direct visualization of the muscle fibers within the motor unit, so the results of the single fiber study were visually verified. We validated the results of the Hamilton-Wright model by extracting the muscle fiber placement information and modeling a single fiber study with MATLAB. Both models showed very good correlation between the actual fiber density and the simulated clinical study. 49 Effects of High-Pass Filtering on MUAP Metrics Reprinted with permission from Muscle & Nerve 40:1008-1011 50 EFFECTS OF HIGH_PASS FILTERING ON MUAP METRICS ..... "..-.. _ .. ", ............. _""'_ ... ""- , • , • " ."" .- " ", ,' "--*. - , " - "", .. . . "'''' . "., ...... ''''-' ' ~ ~ . ". ,"'- ~ .. '" ' ,"' " '" M"", w.'- ~~ ..... ex"'" ''-''''" Al' ' ,,~,,~~ • o ,"",,,-, .~ "'- " '>""~'" "- >", ' '" w..'"~ I~ ,,- , ,"~ " '" f""', '"""1 1,," '., .'.1 \11'." . ,~~ ,m" .0-,-",,,-, • ,,, ... J., .......... ~ . Lo., .. ~. """." .~ "'" " ' <Ok, ~," ........ , 'y . ~ " .. ) 'V'" I, ..,.". ,. " , ",', ,""".' "u, ,,<>.'" "''' ...... _"U_'" ,, ~ ,,~_ U , ' , " . ,,'" '" , ~"' ,, ,", "".c, • .......". ....... ",,,to ,, ~ l>­L- ' l '~ i" \1-.:, ,, ' " ~-jH'" 10.""'_>,.",,' \< ":' ,~ " " "'" "' • • ".~,,-,-. • V~.~ ~· j~ ~- ' ''' ' ~'""",,'" " ~ ' ~,', .. ' ,~, ~,,~ '""" L,J'-<" ,"M· ,,,",,,, . k ..... ·. Ii~,"". ,~ _ I" ,'~ '~' ''~ « ,._ ,~ ,_., ~",_i ., ~ ,~, .'...-" I ' , ~ " '-,,",,"''':' w,,""~ ,,~ , f'-' _.·U " ,~ ". ~ 'J ) ~ ",-" . ,~H_, ~ ..... "C.-, '-" " ~" . .. ' V~· " C' " C' ,_ "" _""" ",~ " " " (, I "-:' ,~, b", ' ''''',,'' '" ,,.. '" "." ~k , c, om'>!' " .. ,~ ~ , "'" ~, . , . .. ", . • ."r ~ "" ," b , ' ''~ ~ "'" v.>I~ .. ' ,,, ~ __ "' ""-'T*' _ ,-,._, " '. _..-.,.. 7"y . ... . ~ ___ .---_ ",",. ,~_ .. '.-" ,~ _ · ___ 0·_"" .... · " .. .. " . , .~ ,, _ ~ _._ ~ .... "'~ .~ " """ 'X," ~ . ' "' '''' ' "" _ _ ~ . '.-.•• ""x, " .. , - ~ -oc'", " __ "' '' "·.. .- ..,k -·c-. '" .,·. ,. .". ~"'.' "--•.. ,-.. --.-p-., .<- .. " '" ,,", '._-- , . • c~ ~, ~" , "~ . , .... " ",,,ko~ Lco"", " . ~ •• .-. ~"" ",,"~~ ~"' , ... , . ~ " U, ~ ~,., . 00" ~~,, ~ , •. ,=~~." .'.1 ,10 , --,,, 00, ce ,,,* c~ J -'< Lr~, ,, ,~ ,, '~ ~ _~ lA L ' _ " ,-, ~" •. ~'''''' ,o •. ~ LA,J ...... I, ' " C~ ,~ "" L." • ,, ~I,. ' ,, ~. ~" "j" " '''' ,,,,,.,, ". N,.· "·"c or....,.., ,,,C 01~' ''','' ~~ ~"' ," . w, ~" , ..... , •• ". ~" • ,u"" , .... . , ~'" ~ "'k , ~ . " . "' ,',,",' , · .I "' ." '~."" " O1U_'" "" '''~ <>.-" "O~ ..... " "'~,_. ~" " ·N··; ,,',, 'K ,,,"'. , " n ... " \ ." ,, ,' ,,~. .;," ,", , ~~ """ '~ " " "' 6 '"",,', M ~ "',, •. ' ," ~'c ... ",J"· ~ . ,,~ ~ ". , .... ,:_"-, ,_ ,, n>",_ " CO , , ~ . ". '" 00, ,"., -" . .. , ~ . "" " . \" c" _ .. ,.,.". W~ "" ·,, .- ~ L ",,, , ..... rl " O1U_'" ," ".w.-,, ~ ~'. ~ C, ,v •.. ~ <l.L"" ",.J.', _""A!' ,,~. '''~'''~,,~. '~"k" ~ ,~"l'.' "'· ",·.i.' "" ~ , . ~ ,, " I,,, ,, ,,.. lCONI, '''" ,,,,"co '" , ,~,,, . ,, " ~ ,,,., " '~'M' '"" r. _ " , ,,,,",_. ~ • C~~""" L"" ~~_. :>1 .,~. b'~:' n, ,,, .. , '" " ""., "' .... '<-<, M ,~ , ., . ,,~ _ " "" I ' ,,, '".''''' " ,-,. 'fl '." ~~ ." _,,~ . '" ""-"_"~;' ,l>, ""._~~~ • .;00 ,,' CO. ,,, , " ,. , ,,, " I "'~ .. ~ l'" ' " 'C 51 ,-"". " .. -"" '" '" """" , , , , , "-"" ~ , . , , , - ~. • " ., • ,,,., , , ' «.,,," ,e, ,., ",,,,. ,. ~ ,""0,, ," " <." ".,- '''., • '0- ' ''''' , "." • , """ ,. - '" • c, " '.,""" <. , ,. "., c .' '. " '" " , "'" • '" " • "",,, '" "''' "" ," , , '.CO "" <.,. ,. ' <,.' ".,. ' ,,- No" " . ,,,,, • L " , ""'" " ...-.''-' ",,' ,,' " ." "" ,.-". _., , , ._, .",.- ", . ,.,.-, <,', " " •• "-"0"" '"., ... " .,~" -"" ,, ..., ,', " " ,, • • .. -".'., C • -"',' ''',-" ,., "'" " "-,, ..' ., c.-'-' . .," .".-', " • -"., " ,,"0,-" .. C' , ,. 0" ..- , ., ,"' ''',' • " ''''' '-' '" " • ~. ' '' 'C , "'- '" " ". ""- ,.,,,,,, "." ,. ", ... " ,,- "." " , ' c,'" " "" ~~- ,-"'" ,- ~.- , - "",.""-.. ,", -~, ,, . . ", -~ . ~- . --.~"~ ," ' ~l~" "" ,ri-~ ' ,~ . ~"' ,,',' 'n-, \f{ \ " ,,-,, ""'. ,,, •• ,'- "-l' '' '-'''''' .,"'. " ,' ,,' ,,'­,, ~ , .. ", \16'--" , ~ _"W .. ' • . "6', •. ""'M '" ~ '" •. J"" " ,~ ,UL' ~, "-~," "-" . j, ~'J _ ~" ',,",, H .- " .. r, .. r,.·., ,, ~ , , ' >-;,~' r, ,,,,< '."~ ~ " -, ,,,," ",. ,'~ L,!,'"" ••. ~ " . ~.,~ ~o i>., .. , ~"~,, ",' .,,~ .-.-uol , _, ~"' . " ." . " ' " ,), ~ .-~." , . , "'. " '" , ~. " ' , .- W~ ,n ' n-,~ I",', '~' " " ~ "n" '.'~ , __ ", "" .. ~ '''' ''1 . ,,~ , ""er '"" m, I ~ " , " ..... ,,', ",. ,~" ", "'f(, . ~.;,. ~ . ! ' ~ ~ , ~ "" ~"-, o! ,,-, L ... " <~ " ~ ,o_,,·~ ~L_, , ,' • . , ,~ ;' ... , e "."' ~' , ~ " "~ .• ,. ",. ,,' , t., ~ " ,,~,.,," " '_'"''' n·M' . n , ",.., '.~ .-_~ , " '" ,u-. "-""" ..... k,.", "'-." • .-0 ,,' " ' m 1 '"~'" W ~ o>'H ,.-.i . ,."",.""' ... ,-", . ~. di..",",,, i>. _~l 'A[' , ~' " '' ",,' .... '" r," I, · ... .,." ,' , .. " , ~ , ~,,_, .... 'V:t'1 "~ ",.., _'j" ,• . t:' II ,. '" ,,,,~ , " m ,,-" ,,' ,>." .. , ~"-, U •• " '''''' ,,.e, " ~ " ... ,, •... <_ • . , ,' ~ •• "" " ,' " ",~ "' .. ,,, ~ "~,, ,. " ,' "' .. "~1 ''';'''··· .... .,. !"~ " '" " ' ,'" r,,·., '"0 I~''' '' ;~' '''~ -" <",-;r dM"~ n."_' , ,~ .. '~ " , , ..... ~ ,_, J"" " ,~ "-,-, "",,,,01 "" ~",- , . l,,, ., "'c ' -0 "'PO' ,,,, ... , ,. 11, ' "'" ,n ,.',l-,.", ~l~ "'''' " ••. ~ ~ _ ' "' 6' ,,''''''' '6" ' .• ' r., .... " ,~~ "-', H, ,", ..... ' .~". ~,~ 1'",-_ ,,,,l ~l*"' " I,' ". '·h . ~, ",. ",.. "" ~.~"' W< . ...... " .c; . .... ,," UJ ~. ,, ~ ~ ~ , ", " '--" "'~" .,~,-~" , ~ .,. ' <"" . ',,,' ,,,,,, ". , )~ . ~, .. ,~ " •. ,, ~ . , ~"6" ~ ,.h " ._, _ ,u-. ." .... " ",," ,,' "-', ,I< I ... , .... .n'"" ,_, ~',,", ... ' , . , ,",, " ~" "" "" " ',", T',' .,~., H L , ~,~ '" " ~ " " ~ r~" ''' ' ' '''~'' ' ' ",. "".;- " ~ ~ , ___ ., , ' . ~ M , , ,~._. "' " ._ ,"" , w. ,,' __ , " " . " "f' ".,. ~' ",,.. ~ ~ ", .~." "~ , ,,',,', ',"",,," .,'" ...... , ' . .0 ""'" .. , COL' ~""- ~ ,, j~' ..... , " " "I­" " 'J " , ,' ~,~ , ,', . • ,e'''· '" ,., .. ~ , ~''' .. w -" . , ~.~"" 1,,,," ,_ "" _ 'j"" . • '" ", ~o" .""",. "">"'--" " .~" ~" ....... "~ I ~' ih','" '-<M", ,'w ,~;--",'> .. ; " ,', ,,, . , ,,,~, '~ b W~ ,,,.,. ",., , . ~ ,, ' , ., ,,1, ,, ", \11 ' .'" rm n • . ,_ " ,-, H L .\1.' . "",. W~ 1""'''"'' "'M~ "'~ It _c-,.., ,"" " ..... ' ~" ,._~ '"" ,~" ~ ~ m, \ml> • ., ,,., .. , r~ , " 1 ~,. , ' "" ., ;"~_ ,, "" ., "'" .. ,, · ,m~ L "":", H L .. ,,' ~U"~' , ,~" v ,",,,,," ,,' '""""', " U,,' ~. ,"-~, "-",l n, .1< . "),, r" , ~ " '< ' " ",." , . ~ ", "_,,,-," "., _.~ ,," ~ ,.,~ ·ru-. ,C", '" ,-,<""",,, r" , " -0 ,n '" '' ,." ."'" '" .,., ,, ,1 ", .,. ," " T~ ,~ , . "'~, '" ,_ , ,'~~ ,. " eM . ~ _ ~i ".<" ,." •. ,," ~"-, ,"" ~. , "-~, . --~ "' , ~ I me '.,. ", ~ , " , ."' 01 .,he", ,' " ""~, .... ..-u-. "'- ~ ,,~ _'lU_'" " ,' ''~~ . .-.-uol "" 0'»' K '" " ' ~ , , ~ . ,, " ._" ", I,,, ,, •• , •. ~ ,." ",,", ~'~ .. .-" y" ,.. "~" (.nj " ,,~ . • _ " ""'" .","",,-",' _.~ ., l<' UJ ~. 1'-'."_' w., " ." " '" r"". '~ ,,. " ,,,, Il ,.'''''' ," " ; ~ ,.' ~ " . ~ ", I",', ,6 ' , ' ... , ,,_~ , " .• ~, "" , .",""". ".""-', ., l<' UJ 0" ",,-"-,,' b",," ",~ , ) ' ) 'M m ,h ;",," ~ '"'''' ''- "",, ' ._ '"' , ...... . _,_~ .... I ... , ~ " UJ'. m" ~ , -"_~ ~ '"" .,.". _,," "" I",' e, ,.-., ,," ., ~" ""', H L "".-,. ~ , )"' ~ ~. "". ~, . --"",~ " ~ ,· ,,,, · U L'l" ~" "-~ " "-'" ,,_ " . ~ ... " . ~ " ... ~ ~, , ... , --e ' , .'.1 •• e J ,",,1 , . ,,," ,','" h ~ I ".".. r> N, ''',, .. ,'''''. "" ','"'' '"', ,,', ,"'-" " .. , ,. 52 , -, , !" - ! -i ! i Ilrl " i • i • H• &i· f .~ hd ~ ) • ~ -,, , • I , ., . -~I,'""~,.I n!, . ---jI- - --jrI -- ,'~ !. h'-.-,s',­" ,; f ,, • "1 - .~ o! itL l '- , I , ~ t !: ~ i I ",~ "1__ ' _',' (-.•- ,"~ -I_ ";'_ ~-H _<r~"l"~~ ~ ;~~~; , _}~ ~ : "!i~ ~~ ~~.~ ~i ; ~~lE~:~. ii ,>. ~«l~'"" . ~ "' -~"~, '_ , - f.., ,­._ . d, ~ ~ , -~-, i~ '- ~ ~-1~',-- o, 'y-~'· f· c!'"' -'' .~. ~"c <+ ,< ~~ i" ~ ~ $;i ,[P :~ : !; '',, ,~.-~ ' ,c~ ".~'; ~_~ -l-I~'l - "-' ·,~" :'.' ~]" .".,_ '' ''' l' ~__ ' ,~_", ••'" ~ r -" , ,I !;~' ~ -'.~(,'.,~;r~_~ "I' ., . , =,. - /~""-' jd~~ ;f~ ~~ ~ ; ~i.i ,~~~,-,~"--~ . ' - ' - ~,-}, ,~ ~~ ' . I ;'! ; ~~ ,;, . ~'. ~ *._,r .. ) ' "'~'<""~ ,; c' f~'f. ~ ~ X' 1 ; ) .,~~ H~1:1. ~ ~ -dq '1 1; ~ ; ';1A: '<·I·-~-~';": ;:' '. ;-i ,~, ', , ,-,-~~~;~.":r."~"j·''· ,~-i" ~'""1 1-•! ~" ; , .. c~.~~ ~< - .. ",.-­~ ,~ ..- ~, _ ,, ~ .~'''~d~:I~ ' _ ' j T ~ ~." . ,. " i-" ~ l " ~' j- - - " " -' _' " '- ~ _ "", ~ , r ~, ~ - c " "~"" - " "'" "~_ ! :;'I ~ o ,- ,~ , n , ~~}Pi ~ . >, ; , 1> !~':~ ' . ~~ "" '~~ . ~r ;: ~:H':~l "'[~'1- ~;~ ;;~i'~ : '~'~"-3'-Et~' . , ,, ~- '~~'~ " l_[j- t,~ I: ·· - ·~ l "· } "_-~,d-~~·,· "-• ~l " i + '".~~ '_ ,,j~ ,/.-~-~= ! ," .-,'' ..'.'. '~,'" ~• ~ ~ 'l i;· ~ c c .~';~'A >- ~ ,'~.- l , . ,, ~ , •• r, :" _ :;};\ -,~_ ;!~~ ,,, ,''', 'i .< l' f~~q~ ,o~: n 'llr ! < _" . " . ~ , • • ~, 1. - .' _ ~! , " l<lf , i ~d.~,-J~' ;~, . ~'\' ~ ~" L : . F, _ . . ~ lh g.; ' h nE ;- : ',' ;';;-L - ;:!- ~~~q; tHo: ~ ,r'.'!, ;_~Fi~,.{ ;1~~,,1~ ,!,'.!j"·:"~ ,;0~,I' :~ l.' !' ~~\' :,:",~~;,/.:ti ;' l, ~1l ~ j ,•~ -i " . •• • !: H I J ~ l a \ II " - I . 1"1.; q,,, , • !, .. ! i ! •• ,• .I ! ,, .I , - - - • " - - - - • • , I , • ! f • I I ! f ,• ! 53 • j nj>l:l ; l~ , ., ' " 1 "", 1, ; - ; til i j ~_ : y d; 1 1 , (' I I, -' , 1,•, .'"~ .I '"~ ,\~l''.''1--" ,•". j1"!"• .'' ; f , I ; , , ,• , ! n l : n'~p~ !. ;1 I"l l ; h ; , . " 'l,((" , · ·!< I d ' -- " J ~ Jk . ~; I - I -·'··: ' ~ d . , •'; ,r .I~ ,."".' .0-. ""''.It,,' .l-.;,•t i.'";" t ,,-' , ~ ' .• ' - . ' .. '~ ~ " 54 Temporal Dispersion of the MUAP When undertaking the work for the first study it was unclear what, if any, effect needle placement along the axial length of the muscle had on the MUAP waveform and metrics. Muscle fibers of the same motor unit do not share a uniform radial diameter; rather, there is some small distribution of diameters. Since signal propagation depends, in large part, on the diameter of the muscle fiber there would then be a similar distribution of action potential conduction velocities. In theory, and in computer simulations of muscle, this means that with the electrode nearer the motor end-plate zone (the area where the nerve innervates the individual muscle fibers of the motor unit- neuromuscular junctions) one could expect least temporal dispersion and greatest temporal overlap of individual muscle fiber action potentials making up the MUAP waveform. As the electrode is moved farther away from the motor end-plate zone one could then expect a greater degree of temporal dispersion in arrival of muscle fiber action potentials and change in MUAP waveform reflected in changes in metrics (increased number of turns and phases). Muscle fiber action potential conduction velocities in human skeletal muscle range from 1 m/s to 10 m/s with a mean velocity of approximately 4 m/s (47-50). Thus, at 4 m/s action potentials from two end-plate zones 50 mm apart results in a difference of 12.5 ms. This is certainly sufficient to distinguish two separate muscle fiber action potentials within the MUAP. Additional complexity comes from the concept that conduction velocities are variable along the length of the muscle fiber itself (51,52). For example, with large distances between the motor end-plate zone and the electrode separation of waveform components could reduce MUAP amplitude and increase the number of phases and turns. The hypothesis is that if there is a great degree 55 of variation in muscle fiber action potential propagation velocity a significant difference would exist in the same MUAP of the same motor unit if the recording electrode were placed near to or far from the motor end-plate zone. We determined if large distances between the motor end-plate zone and electrode would be clinically significant, as it might be theoretically possible to induce false positives for neuropathic and myopathic diagnoses by placing the EMG electrode too far distant from the motor end-plate zone (53). Of note, the electromyographer does not know the distribution of end-plate zones and sites of needle electrode insertion are blind with respect to this variable. The study was undertaken using the same two computer simulations as the filter study (Hamilton-Wright & Stashuk, 2005; Karlsson, Hammarberg, & Stålberg, 2003; Stålberg & Karlsson, 2001). The models allowed the creation of different numbers of muscle fiber action potentials, various distributions of muscle fiber diameters and resultant distributions of conduction velocities. Measurements of the resulting MUAPs were performed with the simulated electrode at various distances from a discrete motor end-plate zone. Similar studies were performed in human subjects with both healthy and diseased muscle: diseased muscles from patients with amyotrophic lateral sclerosis was chosen because motor units will be in the process of denervation and reinnervation and a number of newly reinnervated muscle fibers will be atrophic, thus accentuating temporal dispersion. There is a distinction between the motor point, the site where the motor nerve enters the muscle, and the motor end-plate zone, the area where neuromuscular junctions are located. The motor point can be and was determined in the biceps brachii muscle. 56 The motor point can be identified electrophysiologically and the electrode positioned at varying distances away. The motor point is identified by stimulating with a small surface electrode over the belly of the muscle, using moderate current well below the maximal stimulation intensity, in various places until the greatest number of motor units are recruited for the given stimulus intensity (largest muscle twitch). This muscle was chosen because it is a relatively long muscle and anatomical studies indicate that the region in the muscle where motor end-plate zones are situated is relatively restricted. It was found that while the predicted MUAP morphology and metric changes occurred in both the model and the human biceps, the changes rarely statistical significant in the human muscle and no changes would be considered to be clinically significant. 57 Effects of Intramuscular Needle Position on Motor Unit Action Potential Metrics Reprinted with permission from Muscle & Nerve 35: 465-470 58 • ! , ! ! ~ •• • • ~ • ; • ! ! , 0,> " ~ J ', ,' o >, "" ~ '"o j !~ ~ 'I' l ~:, ,, ~; IL ' ,:) ! ~" ' , -" '~r ·~ "~ "i -';l ,· ~" , ' , . ~> . r~ •.• ~, I' , ,,0..-',:" ";,. " ",' j",'.,- ','..; .!.,, 1.-~,,, , -~, :,>, . !" ,,.,-. 1" " i P~'.", '~; "\'' ,, -f'c~' , ,~_-, t.," .,o , " . , , , O ~ ,~ ., _ " t ;. ~~ ~, ,· ~ , .~ . ; ~ " · " ·-l"" . "' T. Jo,_ "._ , "'_'_'"' , ,", ' y _, ;Y "-~~ _ ~ 'i ~ , ~.- ' > " . " .- . '~ "_ " ',~,~" . ., '----"' ,"" , -. ' " NO"",'" . . .~ , ,., ~ , .. , .. " "". ~i ~ " ·~~_ i "- ~", .V ,~:-;-,;:;)~, ;, ~ ~ , ~,. _r,~, · ." • ' ; <, l . ie"" _[ "",- , ; ' 1 ' " , - " " " ', '.; . >, :~ ~j : : 1, : i In n . iI .;, ,.~"' '""",~"".,'~:, " " :~ ~ " ~ ~". " "_' " c ~ ~ , - ~ ~ " ~ '~->" ' ., ~ ­.. , " .. .. -" , ' "' ~", ~~> ' , ~\ .. , dp~. • ; r i ,'• / ..• .'., I ,;.:,,~ S : , ·, ) ;. . . ' ... ,- '-,. .. ' •. ~ o: <. ='," ~ ~ ' ,, " of " _ . , " " < ~ ~: ! ~ I~ ; ' , ""' ;r~; '-~ ' . ~ -" " ., ," " > • • --'~ '~ . ' ~ ' ~ · t i'i'~"- ­- t l-l- ~- l ;" · " , . f" "~ ,'· /'-P; "~ ~1 - ~ , - ~ ; ~ -> "-'-- ~ " ' .' -, ,, '~~,, .'- .,' " ~ . , .,,~ , " 1 - . "i',',l' .f ",'- ~ , ~~ . ,,";. ,, '" _,,. ,''"0' , ,,, ,r·. _ ' . ,~ - -,~ ' >' " '.' , .t'_'~"· " _ ,_ • .; _~. ; ~" ' ~ " ~" O ' " < , ~ ,;~ - ~, '~':" " , , , ~ -,,<",, , - - '~ ".~ "-, ,'" , , ·", '~~·l~ ~ ~I;. ,,~ ,. ~ l' , , , \ ~l :,,"U'; T~ ',,,.,' l ,_V ,, ." ", >' ,,~, c ._ " l, ' , ~" " " '~"~"i ~ ~ " , ,'~ ~7~'- ' , ~ , '" , ( " ~ ' 1 , ~ ~ , ~ ;H ~ ~ ; j ~ " i I " , , '~ ." r ni \; ~­~ _, 1 ,; i !i i .~ t ,~ n, ~ , , o. q .•o•n' o· .0 cz-' _. •• n.~. 'c ~~ .oc. !-;. '0 •"•" •.n~. •~" ~ 59 I ,,, ,,:; ; ,·• ,• ,• ,• , ~ • , ~ .j~ ' . , , ~ -', ~., ~­~. ' 1,", " 'i" ~ ~,,, ~ C . ~"_, - ~~ c "~~"'' ".,c' ~ --· - · ~-·t ~ " ~ : ~ ''',.- ; ~q-~l ~ ~ !~" i u.; E ~ ;~ : ,.,r ; ,~ ;~ ,.- ,,·. o ' o_~, ' ~ ~ . ~r;' - ~ • ~l71~"o" . , _ , . • _ F _ ~' " ' _' _ .. " '" "" "! ~ ' ,," ," - ~ , " , ~ " "' ~. ;" , "'~.' ~,-- '' l , ofll";: _H ,; ~ : ;"~':;-L~", ", . ; .• j_ . ",_ ~ \ . • _,0_ , )._ ,_ , .\"-'1 :,'"~ _,, _c~ "' O ,~'_ ! ,', 'J;~c"~ ~""~-~__r j;; - ~) ~ , ~-" ~,- ' ~',~ ' -.. , .-- ~ " .;;, ' ,, ~ - "r ", ' j~·c ,.'" ,~ -~ tq· ' .~ ',--~3, ,, ,~, {r ~ ·' c. ' ~ ~ ; . q ' '" "~~~ ''.;-~ ;~- -~,d" i c" ''- , ,. ' - ' ~ ' , ~" .• > . ,, - , -, ~ . - -- <" , lli~ __ d ' ~ ' ~ ~, ~ ~F-' d:i F - - ': ~ < ~c " '~ ' ~ ll " , ~ ~'" ~ ." - ~ ' ~:' " r ~ ">~ ~"'' -"~.:.-' ,L ' '.'. "~" ' t~ ' E'i. \H _ ~ ." '' ~~"" '"'~ ~'-\~I~<.'' i~- ~'. ~-'~_; ~P-.~'',I " - . ), ) ~, = : i~: E;f!~; : ~~ ~ !;l:r:i ! !!I:: ~ l : < ~ ~ ,d,: ~H if ~ n; ~;~H~,;­"" E;"'~"';' l~'~;~' :l " . <" ' , t·~ 1~,. :,3 ;~,~ ~;~ ; c,' f ' ~~ ~ i i" .:­'. " ,' , .~' ", -~ ;l."..."-. ~'<,'~. I" ; '~~> '"",~~.. -,". >- '~ ",; 'ii ' , "_)'~"~ "'._, "~ ", .'. j ~~ _,''_. ,o '_ ~ "~1 . ~ ~ " ;- ' ~~ F ~ ~H ;';:;'; n",!'! H:' ~ ~f , , ~ · j"d ~~H~ , ~ l,: n ·, ' t,, ~ t ,· . ~:::. , ,. "" ."7 , ', ~ ~ ~ ' , ~ ~l ~ " ~ ; , ,<->.;' \_ '',''' ,"~,'_ .'. '" ,''~<-~I~,' q' ~ ':; ~ 1' , ~ _ . ; ~~;~ ; r ., ,, ', f ,, ,., , • . ' O __ " ~ T '" ' , , .. '-~., ," ,- '.\rt,i.,,,'";· "" ,,O" "~-' ,'' '' ' ~ , .~ . d ~ ; " '· ; ~ · ,' ,-",,r -' ,_ ,~-_,- _ . , " > ~ f . ';~j-;'C ~"') l l" 1 I · ,.~,,,~" " " t ; . . " ~~ i , ' ~ rl_, ~ r"r"~~~ : c :J;1 ' l'. " ' ~;~'~;'1-i ~ ~~ ; ~~1 1 1 " '- , "~' ~ ' l, _ .. . ,H~h~i . ;, , ;;,t..;\~ ; .' 'i ';' - ',I ~ 7 ,-1 1 , ,~ ' e 'r:"; - ,f r.~,~ ' , ;; ,,~; ~~f ~/~', ,., ~ ; '.~-",.~ . ~" ' > "'1f'c~ tl' <, ,' , - :: ~ H- ~"l:i,;\ : _ ~: : >L~ < ~ H': ~ '-;j:~i [' .~,. ~;· "" '.r'; "~. ",'~;''~,; ~f' ~~"..'";: _:_ "; >.' "'''Jl' -. ~,, ',- " ~' ~ ~"" " - '''' " ,_o __ i-' .<; ~ ~ __ ~ . , O '". ,;:, nt h ' _:;;~ ~ n 1-"5 . £ ~, : ., ~ " ,,",,\ -., . . , , " .. , '- ,?J , +.q.-~~ . r ~ ' r - ;\;',-: ~ , ; ,,;-1 ," ~"~ ',~ ""~"" " "r" ,"j " np '-"". " ~ •'''., ,~. •-- ', ~>.~ , , ~ ~ "'· ',,· ,i ' . -- ; ~ "",-~ "' ~ " l'~ ~ " '·'! ,, ~: · >!' ~·)~ ~ ,'~ \ " _ ~ vrj,~ -- , " '" ."-,, ~"., '~. ~ ~,~--; .. - , . ,. ~ '. , " ) ~ " " '~~" ~ ~ ~ ~' .,',~' r l, ~ " · "1 I '',~~ -~~l" ':-~-;-;fi~ ~_ ~~ , ','~') '' ~- t-', '~'-, ' ,'t'"~.'~ ; r '~'". ~ '1,,"~ " ' " ' ~ f" ~ ~ ~~;-~ ,:;; 1- t ; ,~n ,-~t .~~tr ~ . ,), , ;-,~; ~,- ; f , : , ~ '; H ". ' 7 ~,; • • • ,,,, , ~ ~. , ~" ''"f- , , . ,. ~ . ' " ,', l .• l, ~ ,. " : "' ~ ~ . j ~ :: ~ '~I" , j " ~ \ . - ; ~' -- ~ •. ~ ;1'1- " .' , __ ~~, 'J , 1 " ' ! - ~ "' ~ "-<' ~ ' r E '_' ~ ~'; ' ' , ' "l ~i , ,. ~§,~ ~ B " =' , i~~~j . , j , , _ "" ', _ , ~ _ o " , " " , . !< '; l ,(,o;_ 7"~. f !~ '~n ~p it.' ; ,. ,. , C , T ~ dl f .'" "'e" - ,, " ' ~ ~ ! - ' ~, I ~ ";''-' ' ''o, ' " 1' 0 - " 60 ! 1 I I, \ ! ..i:Ii -.-oeiiI""'i 'I tll lll" I I! , ,", ' I' ,II,' • • '~ ---,, .oiiI",I .1 _ it ! -..;c.,.i ! , " ; 1!!)IiH' 1 i 1,, .. . • I I J • • ,I ,! I • i I "" ~ '; J P h ~tH :" ""'.1 ~ -. -i ~'d ' .. ·H " , ; ~ . · .. ~ i - q~ , i !, j. ,I !, '!­~ , ; I • ;o: .:. ~• !.. i• -'-I" ". 1,1. j a ~ .;, ·~ l l j'~ ,L, :'; - -" 0: . b r l .' , h H: ll · '~· . ~'' ","".:,, l, q,~ , :" . ~~";i ,~ ',p .I r'q ., - ;;: • ; i . ~,"; .;'j' ~ . !l n III "",.,",. " /I ~ ~ ~J ~. ~ ", • , •• • •• , ~ ~~. ~>o -~;.,"" -. ~" :, ' ~ .. ~,-~" i",,'~,~,:" ~, ""1". ~'.".l , ·"",~~.·, ~,,~·~~-_~ ji, :~~"..'-'<,'; "'~- .'•'""'•"-' ''- 1,~ ;".~i' ;]:~~. ~ . ~ ~ .l i~" - , • • - < _ .• -, - . - ~~~~i'. ,~~,i', ;~~.-"i" •~,l~;"i -t~~£'l~ !·~"~,.~";'");~ _"~"i~; '~l._-Uo-~~-;"j"--·,:~."'~~!',' "~~'-;l "'l.~,j,~: ;"r",:'t,''-:' ~~;~.:"e ~,'>.-. ;t._~{•:­~ ~_.d" ",,"-' , --" , ~ l -;: ? , . f C ~ , , , ~ ~ . ~ . " ~ , " • " ·f ~ , ,, ' ,, ;1 ~ ~ , ' ; '.e.' i. I', ' . ' , , j 'C " ~ . ~ ~ : ~ ~', " : " ~ . ! = ,. . ~ ~ J-; ; " ~ - ,~,:~ , ~~ I:~~ ~H;; ~ ; ) -'~~1 L~; · - - j , .,j- - ~~ I ~ ,~ " , ' .~ ~ l , " , ; . ~ . ~l'~; l :- ,"' ,,. .~.,. ""l . .. . · ,.~. o- ., . -~,, -~~ ". ,,' , ," ,-' ,;' .~ ; -; 1 n ~ -:~-'.. -' ~~ 1.. 'I ~, ' ,~ ' • ~ ' I', 'i.. ~ -." -1:, ~ I " " > ' d~ {~, ·u _ , ~ t, · ~" >.,H. t"~~·~,· •-~ ~'~ c"; ·~· .' , . , " ' ; ~. ~ , ' ., .' .. , , > ~I~, ,-- . ~~" ."," , .. ,. . ~ .~ .~ " ., ,,1 H","''' ;) '- ~ ~ .. ; ­,, L ~ • ~ :~ ~ ~ ~ I .r ::,l : . .o , 0 • - ~ ~ , .. ~ c ~ 3 ~ b .,, ~o · ·o, 'I " ~'::­r ; l U ; ' f ~ ~"> I""· ' ~~ ~ -- '\' .... l , ~ ~._ ;i"~ ~ . , .. .o ~ ",," , . ' , , ~ l : --"' '' '," . , ' ,", " - ,, ' . .~ ,. ' , ~ " ~ i f~ " ~~";;~~ ~ ~ ; ., ; , ''-lU ~ '" ~ ~i •, . •. .~ ,". ,- ·.l .~~-, ~,~ , i. F _" . ·_ '., ! ~'~ T ~ _ < ~. - ~ ., + ~~' .,""~' ~ .. ',s~ ' >\ "" ~ " T'?~ ; "' .~ , 7 ~ "o \ " -- l '< ' " ~ ~ ~, -~"~ .,' · "" ·," no _·,l", ' ·'~_ .~, ~ "' 01 . " ~ , .. ' ,"" " ~.l " • '1" 1'"' ~"-"·' -- '" • -i1'; ~"1"~ 2",1i;l~' ~,,i~-.~ I ~" .• "'-~ ~ ~ ,. • ,i I , ~ ! ~ I ! ! ~ • 61 I ,,, ,,:; ; ,·• > , ,• ,• , ~ , ,j ,,,, I I , I i •• , C - .! -b-,­_ " , ' , ~ • • • • ,,,,,,• •-••• , I ! I . !!uii i . ~ , ~-­. ~ . , [t- .,,,. .~ .. - ~ ,! -i ~ .~. , , - , , ! ! I l"'l~ ' ~'I' ,~ ~ . ' , T " ..•' ;~"" . " "j ,t' . , _ . _.~O. ~,~, ~ ~lF- : "" ,~ " ,, ','. 1 ~ 7 " .~, 1. 't · ~_fO t.·- ~.~tl l,'<~. , ,. ,,- , ", • r. .t ·· ~~ , . " ,,_. " ~ l "~ .~ """ , l · ~ )' " ; ,o i ~ ~ ~ I ' ~. --,' ;, < ,-.' ~ '"~~"' ~ " .· · -"" l' ,~--l-~".~~ ~' l.H".:i;r'~; , ,- ,, ~;.,. ~ .. i ,' ,"_ · ' '' I "~·'l, .'~ ~~;''- ~3- < ;l ,~- , 'i.· < ·· ~ ', I P'~ Ti'". " .- ~ c ~ ' . , ,-. , ' . - , ~ c , - ~' -- FI , ~t .< ~ i "~~ " .. -" ~ ~ ; .- 0 •• _ .;. , ., - ,- ;' , ;' .! ~ , .'. f i ' of ~ nt ,,' ~ , - - : ; ~ ~ ~ • , , , ~ ~ d- ~ , •( •~ ~C , < 1,', -~+ :~ , " ' i'-~',,~~"~ ~ '~_li"~"h " '•"- -~1"-,-''" "\~. ,,'"- _, ,,• ! ' ~l_,,~_ ~~ .'·, -: "' l~ ·_" . •1 - . >--. f_ ~t":;~" :l l- >~_r He P,' ," "j , . ,~ ," ,­, , > ," I -l'. "- 1 , < , ,.,, ,• , ,,,,\ ~ i i , !, :',, ," > , " ' . ' .. , ,, .!" .•-' .>, ~- •. ; ~ ; • i! , .1 ;, -.. , ~ .. " ... ~ ... , ~,,~ , ~~,,- .. •• ~ ~ p ~ ~ ~ ' ~Lii'" <, i>~ ' ~ . , " ~ """ .. ~ - .,,) .,;:'"· ~, '~ 0~'i ~.' " . _, ,,. , ,'," n~ };,, ~ ~,~ ;"! ~' !8- ";5 :.~-_i-"~ c. , • ~ ; ~ ,, I i -' ,t < 1 ; ,, , : • , ,, ,< i , ,,, ,, , , ,1 , , ,, 62 ,.,; ,." ",", .. " "". . , ,0: """" ,' ~"'. ' " ~,~ ." """, ,., .. ,." ; ,." , ~ " I' '"' •. ,'~ , ~,, ~I,~ , ... ,,~ ,,>J ~. , __ " ." .-. ',,"~'~' ." lUc ~'''.~ "' I~ ,m,," ~ ,'-' -,,- "" ~ , , '" ,~ ~ __ " " .,,," .. C , ,", ~ "1 ,,- , , ~ . ~, ~ ,., ,. ,'. ,." ;" , ,, ~ ~, .• _,·-.it" ~ ,. ' __ "" ,.,~ '1l'.'" ' .,",w_, " ,n _~ . ~ . '0. ,_., ;Ie,,,,, I," ~ "" ~" ,., . . , ~." '" ""~ ""-.,, , , "' , "~,,~ "~-~',~ k ~ '" ~ . , ~_ , .. w .. ~"'>' ,., ... "",' lL.· '--C" . ~,: . ", ' ,<" " , ~ 0 d " ' . ' .- "" , ,~" , <' ~ . W· ' '. ~ M ' - ;--"""",,",', ,,,,,y. ,. I, I ~ ,~. , "'-~ ,." ,-,. ~,~",,, , ~-~ , ,- , .. "", ~, .. ~" , - f, ~ II, ~, " .,,-c,, .. I~ ".,_~.~ N, k " h ., _" . ~ __ ~ , ~"_- ,b __ , .. , ' ,,, ,., ~,b'.-, ~£' ~ k~_' ~ .w, ,,~,. , ., "'_'>' ~,_., >,,,, .-,-r, ., -, , ~ 0 - ~'"-<'," , ~ ~.. " .,"" ~ Tho ., ' " " .", , ', , "",, . .-!e. , .. ' ~~ I w, _'d ' ''1' ' - '' " ''' '" - ~j,,, •• 1 '". ,,', d ,,,. , " ."- ~ ,_ '"~. ,,,., ~' A ~ ,,>, ,."' •. ~. _ ~ ~ . lo-" .. .. -~ , To, "~ • . , .. " ... " ' C " "; ' ~' ' / ~*, n " ,,',. _c .... "I , ,. ,.', ,,., 1,' ,0,.,,'''''--' , ' ",.,~ ,.,' , ~l 'AJ" ~~_ " .,,~ ... ~ ,,~ ,~ ... ~ ~ " ._ ~_" _ ""., " ~~" , ,; <, ."" > " . ~ "". "' ~ ; .. ,,' , , wd ' ''~., ' " , .,,,. ,,., , ~d , e"M~' ; ' d _ ......... ,. u_ ... ~" _ ~. , . .... oL .. ,,-,,,, ,,', .,,~ " .. ~,,"'~ • ... <" " , ~ ., . " .- ,,"f,,,," " .""" .. ,., ,"~~ , ,~ ,,-,C, .. , .•. _""' ...... .. ' ~"'" o_,·w __ "" ,,_ , ~, ' m.· . ... ' ,',Y ' " , ,' "'~ ,,~ , ,:;,." , , - ,, ~ ., . , --i' • I ~ , , . , ..... ,. , .< ."""",, . • L.~~, 1'-'" 'U' ,_, w,," "' 0 ',-+"" '0'''''' . w •. ~ .. -. ,-'"' , . , ~ ~." " e ,,,.~ . ~ " .. . 1 ,"' ~ '''. ' ; , ,~_" 11 , ' , " ' " " ,. , .•. ' ~,. , J,","' _ '"' '",''' L, , ..... ",. ~ .... , ,~ _,' "~'. " . ' Jc" "~'., ,_ , ' k . ,,. ,,~<,,," .~~ .- ~ , , ,, ... ,,,, m ~ ,, ~- ,., , __ . .J . ,_ .'_ ..... ". ~~ .. , .. ,~ ~. ,_ ,."p-,"" .. ~,,,, ,, ," "~~"","'" ~ " ' ., I. ~ , I , ' ,'u ,", , .,, _ ,. , ,,, ' ,., -,." .. , ., . ,, ~ ,' , ,,,," f"-'''"' ' ~_' ,--__ ._~. ,'_ ,h ' ..... ,,­.'."", .- . ",,, " - 0 ·",. " fo . n ' ~ eM_ ' " .,.~ . h ', " " " , . ,; ~ d" .. " ,;, ."'-W T' -" " , . , ,,-, . • . ~ .. ~. , , ,," •• ,', ".~ 1.' ' .£ ,>"-"-_, , -~'" ' "", .. ,- ' I . '-" ~," " "' ' .• ",,, ,",. ~ ,, " ... '-'" """ 'C."-'" •.'•' . ." ·h .C. '''~ . l) .,,~' .. ' d~, ". n L~." . '",,'d , ,., ,,,,.­.,,~, ' .>1. " ~'.,","" "" , "" ~. "'" ~'w _ ,., .", d~ ,,~, .. ,. , ~ "'-.1> ',m' ,~. -~ . ,." ., '" m, I ~ T' ,'~ ",,'" " w',C; "., , ,,. ~ ,"'.~~ "'~_' " .,,,,,. ,"'c ,,",",k'" ~ ,. • Co'- "' N I~ ~,., ,~ d,.." ,,_~. , ~ ,"" ,- .• 1, I ~ ~ ' .. . ,,' ri ,--;, . 'I .' ;"~" ' " ",;-, ; ~ " "-",,, ~~, "'"". _ ~ ,,_, , " ".~ ". ,",- ,~.,_ k _ '." , 'CC· " "';, k ",,", ,,, ., "'" ","" I ,~,' .• -I~ ",," I , ,",I"' - ,,~.. ., . """"m» ~ ." ~" , '.'. "" 'A', JC,"_ ,~_",. "_' __ " " ' ••.. __ , ' , - ; -m ~; - " " , ' "''' ' ~,., ' .- ",," "'" ~ ".M • • ' " , d_ "" 1"'-. ' ''~ . "." '" . ""'.,~ r ... y,,,,,,,. "' ., ~, ' ><"' ~ ., ., ", ,,,,. 'm -" .. , ~""' " ......... 'Co·'. ," , -, ,'c.' '" ~ ~, , . I to' , ., - •. '" "" , ., ,C i-,. ~. , 1 ",",,', ,_~ . .' ," """" '''U A'' ,'_ " , ~ ~ ". ,., __ , ~ " ".' ,., '" ~ , .. "'" ... ~ " . ~ ;, ' '',' ,',-­.. ,-" . " , .•. , I • . "" ,,-, I,' , ', ' ~, .. ,. , "',., ... " . '" _~ ~, ..... ~. ,~, M ."-", ",", "' ~" " ., '" ~ ' f''''' ,,~ . ; .",.' (.~. , . m .• ,.,­'" . ""~" " " " , ,,,,, ~,, ... - , . ~ , ',, __ i"'M" ·'" ,1,,..', ~, " i. ,_'_J. ..... ~ . " . " ~., '-<_'~ "" ,,", i., ,~ . ~ ~ ... m·.·.. ' \11 1,,· ,.". , ~ dk' ,~ .• , .,, ~ " .,,' , , d '" .... 'or .A' , ... ,."...-~ ,.~ .. "'_ ~, .~ ... c, _~ »,," C ." ..... , . (. - f ~~ ·"" ,,·. ,, · '\11 1 ", . ,~ , .. . "<"'~ -i-I"" "u ,' , "~,, .,. ,,-,,,, ~ " u,d " , ~ ,,' 'o~, ,~ -" ~_, .• -" "' ~ .... _'_ ~ ". ~ • '0 _ e ~ """ ' . "T>, "''<' ,. ,.". , ~ ,.,.~ " , ~ ~ .;. ,'¥> ,>--,", ,tft·, -,·" n ' '' ~' ' ", wi -I"" "'- ,,_. , ., I, ,,~ .• , I , f""' ''''' "". ", . ., '" >C" , ... ~, > .• . "~ '1l'.'>h. ,.H , ... ,."", " . "...-. . , ~ .. 0 I~ ,,',n),,' ~ f"~" " ," , ~. , _ ~, ~ ,~,, __ I. ~ ,.In" () - ' ~ ,. -,~ .'" -I~ . ~ ~ " . " "-- '-<_ ~ "" A~' ,~ . • ... >"_'"'- ,_, ,>"-'" I~ " . ,'~ , ' ",",', ~ '" ,,' . " ,., , ,,~ ~ ," ,~ 'M ' ."'-W .... , ' , " ,, ~ ., "_' .~ ",~ _ -.."., ., ,_,. , ." ,,_, ~" ,~~_'~'" ",>J ,_, ~. ,'_ " , ., '" ,;;~ , ~" , ., " .... d '" .", ~ ,~. , I __ ~ •. , ,' rM'" ".~ '" ~" . " '. ,,,,-,,,' '" , " .. .. "_~." , ~ Uc ' 1 •. " _'.p ,.,,,."", " "C, 63 ~1, ~_> •. ,'~•. , , . . ~ ~ ~ " ':'- ',; n, ~ ~ ,~ "< ,; " ~~ iL '"<, " .' ; ~ , " ~ i , ~ I • , p "": ~ ,. I ~ ., ,!" , .• ~ , )" ~ I < c B;; , ' .• ~ , , " i J C : l d , , <1 • ' -, ," .1 ' , , ' ; .r, ,' "; ;i l '~ ' . . - .. - ' > , atI~'! " "' '" . , m ~ . !i ; .. ~ . ~ . ./ V ; t 'lU Ii 1; "_~,.' ,"" i•" "', .,..' T. "~ .,~ 1.c ",. , .' _t "" 1 ;,, _ ~ ' ]-.' , " ,..i.-, '. " .\,'. •, , ; f "' .,. )-' ." ' , - , :lc ." " " ' .. ' ~ 1 ~" "j ~~, ~ ~ 'I' . J"" ", " "," , q~ ' ,;: 0 ~ . ; i ' ,c-' ". ' ",-2.' .. -., '~ ~,. ' ·I .'l"' ~,' " : ii l :'~ , .~ l~ ,~,:';. : tj j! ~, ! ," . ' ~ . " ,. ' ~J~'I1 • •i ~"~ -.(' ,"·. r;;,,,."'".- ",,'.-,. ,l_" ,' "I--"' - "·-'f .-~I lq"-; ~" L, .,,' >I •:-,,. 1 ,,;:', " ,10 ·'>' · ' , ,! >1 ,1' - - , .' I I . o j , ~ " " ' __ ' - ' . -" "!i r' , .''', ,· 1 . ~,.' _• ~' '''.' "''' .'i~i · , ,"' f -' ~" "~',;.", O" " ' J"'I. i,1~""~".! L' I ~ JJ ~~~:; . " ~i J" _"c<>='=. , .,> ;,,_ . ... ' --'j "oC ' , 1 ° 1,',"<"0 '.. ,' '_ .,.', '"_. ..,. :,' =", C ,.l "e ,.,." >, c.•. "d ~. ;'_~ . ,·,<"' ,•' ! ~ .',iLi ,:; ! 10 10 }"W;d W ~;;1!} ~~m ~ ,. . " !.3~ ~1j "'-~ g ;' i ~~ '> .;~ : ~~ ~(j ," ; ~j4P ' h ";- ~c;il " " " -""" '1'---; c i .' C· ' I ~ ~.j '" ""i . " " t- ~J ' i~~' I ~" --. ",-., . ' .. . , _.' ~ " • • > ","1 ·' I " ~, ' s",,'" J '~ ' -" ', ' ~J " ; '~~~' - ~~ " "1 '. ~ , ', - ' .-~. , _ " ).,< . ~ ·, ,'!~~ " ~ ·i , "' 1 1 , \-: 1 , ;>~~, ;",;,< ,~, , ~ ~ -.;t / .~ ' :-. " ' ~ "-) "J · · ' t"~",~ ;. l~' . ~ ' ., '_; ;f .. u _ . ! , ::~ I, ~,- ,,; ' ; ", ,: -~ \ ! ,F ~ ~ -, ~ J~ ,- ," • ~""f ' ,".."; ,''~ "- 'l ,_. ~ ".' "." .~ '' ~_ ~-._ ~, r', , _,- j '_ " . ", ,~~" _ , " - ".- ~ . ,, " " ..... " o ~_;" , ~_~. - - . > ~ .-" '1' 1'," ~ ~' . ~~ ., . .. ~._. _ . '.,' , j- ., .. ~, "" ' -1 o}:; , i , ~ ~ ~ ,'i~-7.j·_' 1 ." I!L~; l~~ ' ~ ~ l ~l~ \ 1; ~ t~ "£ ~ :? ~L~ j ) n~'i ~ ~ \ ~ ~ 'l, '- ,,l· '~ ""'~f, ~ 'j "" " ,j ":',,~'; _~- "~, ·..;,~ ~__ ~~." _" ,. ~"t, ":~~ .~~~""~","· ~.t' ~f_ :;>'c :"~d'~,."' .< I " ~ ' > " : "", " ,.j "' :'I~. ' . "~; , . , ' . 'i( ~ " - ·; ,~ j:''' ~ ' l ' jii :,-', ~, ~~ j'\'l~i '~'~ ;i~; i ~\ ~ ~~ i ' " J-,;i' ~ , ~l ; ~~:.,~,~ ( .3' ~ k ,'H:~ " H ; " l' l "d , ~ q ~ \i ! . ' ; !\ ~ ,] ! , '~_'.J~ " ~ "l,' ~ ,~t',~ ~ ,~"~, i. ;, ~~ ,',_ , , ," ~ ,,,• , ,l ,,: j i , 64 Needle Electrode Selection A variety of needle electrodes are available for clinical use for an EMG study, and among the concentric needle design there are two sizes available; a standard size (described earlier) and a smaller pediatric sized needle. The difference in size of the needle is reflected in the uptake area of the active electrodes. The standard electrode has a recording surface of 0.07 mm2 and the pediatric a surface of 0.03 mm2. We sought to determine if there was a clinically significant difference in the MUAPs recorded from the two different sized electrodes (54). Another reason for this study is that the smaller electrodes can be used in place of single fiber electrodes for measurement of neuromuscular jitter. Verifying the use of the smaller electrode for multi-MUAP analysis can expand the utility of a single electrode for complex cases. Further, the smaller electrode is more comfortable for the patient. This question was first investigated using the Stålberg model and then in the biceps brachii, first dorsal interosseous, and anterior tibialis muscles of several volunteers, as these are commonly studied muscles in the clinical setting. To add robustness to the findings, subjects with both healthy and diseased muscle were studied. In this portion of the study in muscle, efforts were made to ensure that the same region of the muscle was studies with each needle in each subject. An attempt was made to insert both electrodes together to ensure recording from the same region of the muscle but this did not allow individual adjustment of the two electrodes to optimize recording features. There were some issues using the model-which ultimately led to exclusion of this portion of the data from the published paper. The simulated electrodes were not created with the intention to precisely mimic the standard and pediatric sized needle 65 electrodes; rather, they were simply "large uptake" and "small uptake" area electrodes. We went forward and completed the simulated study with these different electrodes and found significant differences in duration and area, though amplitude, turns, and phases were statistically similar (Table 2.1). Table 2.1: Results of needle size comparison using the Stålberg model. Results are similar for metrics of amplitude, turns and phases; while results for duration and area are markedly different. When differences were found in biologic muscle the differences did not match those found in the model. Comparison of Simulated Needle Metrics Mean (SD) Mean (SD) P-value Amplitude (μV) 407 (281) 433 (263) 0.505 Duration (ms) 4.51 (1.37) 5.90 (1.7) < 0.001 Area (μVms) 294 (175) 403 (202) < 0.001 Turns 4.75 (1.8) 4.71 (1.72) 0.873 Phases 3.94 (1.65) 3.67 (1.59) 0.244 Small Uptake Large Uptake 66 Comparison of Standard and Pediatric Size Concentric Needle EMG Electrodes Reprinted with permission from Clinical Neurophysiology 118: 1162-1165 67 ,~ i • 'j : i ; q , , , , , , 1 , , ,,; • I • ,,, " •~ '. tId Id 1 j t . " ' I c :" i e , " , " ; il , ~ ~ , t "" , " g l: " ~ •, ~ , , E ~i,p>j," r .... ,". _:', . ",,·f~. • • ,, h ~ - ~-. ~ ~ " .. , ~ '. " " '. , " ,, ~ "'- ' ..... ,, ' ' . "..." ; ;... ~,~I "~ n. ,; ',,;", 1' , " ' -" ,," " • • • < • ,. ; ' .. ~ , ' . t' ~," ~ ­',', , .' ,, ' H it ; ' ~ ~ ' ,,' "-'1' r '- o ,t~r~: -':C ' f'-· c'· ,;__ .,~-; _ .. ~ r , ~ " t -' - -, - - '< c 1! ~[ 1 ~ i ~ ?A ~ l~ _,7 r, ; H ~ ~} ;~ "," ~. ,~ '" , r !," "l~.,,,. ~~ "•, 1.F ,-. ';_ , " "nrrA :' ? -.. , :- .. '_ "" .. H.. ,f,, ', ;n' ~.. "~"?:-~" t "· -"~-~ r1:; ' .~·~ ,~ ~.·~ ~··~ ··j,t· ·; :<~·:·- ,,C o, , rl"j ' ~" " ',," . i; ~ it i i;f ;~~~? V , . ~ .. , - ~" . .. . , ' c ' " " ..~ ~'. 'r"_ . ,, <. ~ -" -~ .,. '~_r !F ~:~,~~~c ,;:.,;l F' ;!-"i ~-;cf .~. ~ ,._ ,_ ~" ..."• '~ T. ·)i. I..· ,,"-:-, _ ,, ,!>~ '_• •_ ". '" 1r "":~; ,'.", c'; ! ,- '. ,~.;r,ti- ~~',:·II,· · -~ ~! ," O'l _ ,. '> . c ~ " "j i._:..~' 1,, 'i- ,"t~ ,-,C.t..' "J ~ ! " ,, ,, ,\ , ce-f,,--~_."J -,'_-"~,.'-"- ,"_ '-­! d; 'i~ '; JP.I'-'F:< , I.'~_. tct- -",:'f,_""i,,~~~'''.~"''_ • P ~!., - " ' , j _ . ".tli t ' ! ~" ' ~~j: ~ ', I ','~' ''trdo"t ! . ? ; k i'f'1 I· . , - -- I . ,• ~t ,;- l' ~ p_< ",~" " .H,,';0~>~,.;, , ~, -,, ",_ "• ,,'-,,..;-,.,"j- \"'' ,'"~~;d, u,,, " < . - , ., ,·,'- >. ~ _l~I!i-:- <,-~ • , " L t · • , ," " - t' , " t - -__ 'r" 01 ,- -- -' ' ,- "1-" ' l , i • !. !;:P' i.! :.. .: ~ i j - " l}4i U i " 7 ,'H , , , ' , ~~ d i ~ ,• ' ; , ; , , i i ,~ 1 ,• ,: ,• t , ,­~ i i: • • ," • ,• ,,,• ,, , , o , ..', " , " • 0 , .•• -,,-"g l§cio G- -~ o"~ .C "0f,r ~ - -2&.. H 2.g !j; :r, ' ,;J - , , i ,,,• ,• • ; , i: ,I. e 68 ,,, • 1j , -J V , , , j ,• . -', . .' , ,• • , ,• • ChI • • • •• -' <: • , • ; , .. ",_ ... - . ~ .H • • •• • • ,- "_ ,.,p , .. " "" , ,• ,~ 1 69 .. , ....... .·'' .'c'·'' ',b''"'c.' s..-~, ~ ."" ~ 70 ' _ .~ ' . ... , .~ ' 0<" ___ .. · _ . ' · ................. d ,"'_ ..... _" ~ -,'~ ...... -_ ........ - ~ -'-' ' ." . ~._ , ' J. •• ",, ~ ' ,~d ,-"', •.• '" , " ~ ' ' o ~" ,. , .•• .• _, .' " .'.' ,., ., , ,,.,; • -<""",; •• '- -.- -- - ~ • • " -, .. , ... - . ~. ."'""". , " , . _, C~, , .. , . , ' ,h, '''-', '. •., . .....'. ...- ,.-.". . '.... _,"" . " ._<,.- -"."-~•".. --,_ .., ~"".,. " ,_,."_ ,, .. ~u,. _ ......... "... ... " " ---., , .'---.... , , ..... ~ "' _ . , _ _ ~ .". "" ~h'O ~ ....... ....... " . ,," .C " •., . ~' ; .'-' .~ .. ,"," • . -" >0 ' " ,- ....... -... ,-.. , , , . . '.,;, _ •.. ". ..... .,., ,..", '---" ~ .. -.. - ----, _. .. -.- " . 71 Optimizing Acquisition Time The algorithms used to detect MUAPs in QEMG studies require that many discharges of the same MUAP be collected and averaged in order to ensure a fair representation of the isolated MUAP and distinguish it reliably from other MUAPs (40). As discussed earlier, the major contributions to the MUAP come from muscle fibers of the motor unit within 500 μm of the electrode. Thus, very small movements away from or toward the closest fibers may cause large changes in the MUAP morphology. Such movements come from respiratory and cardiac movements of both the patient and the operator, from physiologic tremor of the operator, and hysteresis of muscle tissue displaced by initial positioning of the electrode. Different QEMG algorithms collect data for varying amounts of time, 10-30 s. This study was undertaken to determine an optimal time for the collection of EMG data, balancing the need to collect sufficient MUAPs for a stable average with most noise removed and the likelihood of needle movement within the muscle that could change the MUAP morphology (55). Thirty seconds of EMG data was collected in human volunteers using the DQEMG algorithm, commercially available (Neuroscan; Compumedics, Ltd., San Antonio, Texas) and the raw data were exported and analyzed in MATLAB (The MathWorks, Natick, Massachusetts). Two types of interference patterns were collected, simple with one MUAP and complex with two or more MUAPs. For the simple interference patterns, data were exported for analysis with MATLAB. A custom waveform marking algorithm was created. It was determined at what time each MUAP fired and marked the metrics of each individual MUAP. Great care was taken during collection of these epochs of EMG data to reduce noise from any 72 source. This allowed for relatively good measurement on non-averaged MUAPs. Peak to peak amplitude was very straightforward to measure. Duration was the most difficult metric to determine. The difficulty arises due to the uncertainty of exactly when an excursion from baseline represents the beginning of an MUAP versus being part of the background noise, and when the waveform merges back to the baseline at the end of the MUAP (12). A threshold was set working forward through the signal using a moving average. When the threshold was met the algorithm would then work backward to determine the appropriate takeoff of the MUAP. A similar method was used to determine the termination of the MUAP. Once duration was determined the other metrics could be measured. Area was measured by a simple rectified numerical integration. Turns and phases were measured using a numerical derivative that provided sign changes. Once the sign of the slope changed, either from the beginning of the single MUAP or from the previous sign change, the algorithm examined the original waveform to determine if the appropriate turn or phase amplitude threshold was met. A threshold for numbers of turns and phases was set at a displacement of the signal of greater than or equal to 25 μV. If the turn or phase did not meet the amplitude criterion it was ignored and the algorithm continued to search from the last true turn or phase in order not to miss a legitimate turn or phase because of some small noise in the signal. Metrics were measured for each firing of an MUAP (Figure 2.2) and for each contraction length (2, 5, 10, 15, and 30 s) it was determined what effect the changes in each metric over time would have on the averaged MUAP (Figure 2.3). Since not all algorithms choose the same MUAPs from the train for averaging the effect was measured in several different ways: using a running average, averaging all, and averaging randomly 73 Figure 2.2: Example of measuring metrics of motor unit action potential trains- amplitude is shown here. Each MUAP is measured for all metrics. The amplitude of each MUAP, lower graph, is represented by a single point on the upper graph. Minor fluctuations in the amplitude can be seen throughout. Near the end of the 30 s epoch the amplitude begins to diminish visibly, indicating that the active surface of the electrode was moving away from the nearest fibers of the motor unit. 74 Figure 2.3: Representation of changes in MUAP amplitude, using a line fit of the data, of several MUAP trains at various collection times. At each time point after 5 s the range of percentage change spreads out. This figure includes only 12 MUAP trains for illustrative purposes. selected MUAPs evenly spaced throughout the train. There were differences in the results of the various averaging methods for the longer collection epochs (15 and 30 s), but at 10 s or less all methods produced results which were roughly the same. For complex interference patterns, those which contain two or more MUAP trains, these were cut at specific lengths (2, 5, 10, 15 and 30 s) to simulate having collected data for that specific amount of time. These were then reintroduced to the decomposition algorithm as if each were an independent study. 75 Optimizing Acquisition Time in Quantitative Electromyography Reprinted with permission from Muscle & Nerve 40: 371-373 76 1 1 e • j , l • •-w •o .~ ,'.. o. 00 ~~ '0 ."-. _0 o.~. ,, I • ! ,1 j I < ~ ,,, , ,) , ,I , , ,• ,, ,, it .", ,i " ,, , , ~ , ~ ,! 1 • " - ~.\' ~~ · ' >l"' ." l " '~! ~ _,"" • • 0 _.~ •• > .~ > _ ~ .~., ~,,,~~. ;~, . <~'~.;' ~ " ".~ ," ., ~l ·~ t ,~ :<"~"'~ ':'~I'~' :i'" 'v~ .~ ,. ,~ . " .,1,." ,'" , ,' ., ~.-~,; . < ","" ".,"~, .• ~-.U ", i'.~._ '~ _ ~~ l') ~" ' , . ~~ i , '.' ,~~, l ., - " : ~' , :t " ·c~ ~ ~, ~ ; r, " · ~l ~ ;q :2, ~ ; ·>~'~i~.;. ' ; .