Automatic analysis of EMG during clonus

Automatic analysis of EMG during clonus

Chaithanya K. Mummidisettya,b, Jorge Bohorquezb, and Christine K Thomasa,c

Chaithanya K. Mummidisetty: [email protected]; Jorge Bohorquez: [email protected]; Christine KThomas: [email protected] Miami Project to Cure Paralysis, University of Miami MILLER School of Medicine, 1095 NW14th Terrace, R48, Miami, FL, 33136 USAbDepartment of Biomedical Engineering, University of Miami, P.O. Box 248294, Coral Gables, FL,33124 USAcDepartment of Neurological Surgery and Department of Physiology and Biophysics, University ofMiami MILLER School of Medicine, 1095 NW 14th Terrace, R48, Miami, FL, 33136 USA

AbstractClonus can disrupt daily activities after spinal cord injury. Here an algorithm was developed toautomatically detect contractions during clonus in 24-hour electromyographic (EMG) records.Filters were created by non-linearly scaling a Mother (Morlet) wavelet to envelope the EMG usingdifferent frequency bands. The envelope for the intermediate band followed the EMG best (74.8–193.9 Hz). Threshold and time constraints were used to reduce the envelope peaks to one percontraction. Energy in the EMG was measured 50 ms either side of each envelope (contraction)peak. Energy values at 5 % and 95 % maximal defined EMG start and end time, respectively. Thealgorithm was as good as a person at identifying contractions during clonus (p = 0.946, n=31spasms, 7 subjects with cervical spinal cord injury), and marking start and end times to determineclonus frequency (intra class correlation coefficient, α: 0.949), contraction intensity using rootmean square EMG (α: 0.997) and EMG duration (α: 0.852). On average the algorithm was 574times faster than manual analysis performed independently by two people (p≤ 0.001). Thisalgorithm is an important tool for characterization of clonus in long-term EMG records.

Keywordsspinal cord injury; muscle spasm; clonus; wavelet analysis; surface EMG

1. IntroductionClonus is one kind of involuntary muscle contraction (spasm) often seen in musclesparalyzed by spinal cord injury (SCI; Beres-Jones et al., 2003; Little et al., 1989; Wallace etal., 2005). It involves repetitive contractions followed by periods of relative muscle silence(Cook, 1967; Dimitrijevic et al., 1980; Rack et al., 1984; Walsh, 1976). Some individualsdescribe clonus as manageable, whereas others consider it extremely distracting because it

© 2011 Elsevier B.V. All rights reserved.Correspondence: Christine K. Thomas, PhD, The Miami Project to Cure Paralysis, University of Miami MILLER School of Medicine,1095 NW 14th Terrace, R48, Miami, FL 33136, Phone: (305) 243-7109, Fax: (305) 243-3913, [email protected]'s Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to ourcustomers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review ofthe resulting proof before it is published in its final citable form. Please note that during the production process errors may bediscovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

NIH Public AccessAuthor ManuscriptJ Neurosci Methods. Author manuscript; available in PMC 2013 February 15.

Published in final edited form as:J Neurosci Methods. 2012 February 15; 204(1): 35–43. doi:10.1016/j.jneumeth.2011.10.017.

NIH

-PA Author Manuscript

NIH


NIH


interferes with daily activities (Adams and Hicks, 2005; Little et al., 1989; Sheean, 2002).Only a few studies have focused on quantifying clonus in the controlled environment of alaboratory or a clinic by measuring overall clonus duration, contraction frequency orduration (Dimitrijevic et al., 1980; Iansek, 1984; Rack et al., 1984; Rossi et al., 1990;Wallace et al., 2005; Walsh, 1976). Whether these observations provide a representativeview of the clonus that occurs throughout the day in muscles paralyzed by SCI is unclear.Furthermore, the magnitude of the EMG has not been evaluated systematically in theseprevious studies but could provide valuable information about the intensity of the musclecontractions. The prevalence of clonus is also unknown.

We have made long-term (24-hour) electromyographic (EMG) recordings from paralyzedleg muscles to quantify the prevalence and characteristics of involuntary muscle contractionsafter cervical SCI. For the clonus identified in these records, it is possible to measure themagnitude and the duration of the EMG in each contraction, the frequency of thecontractions, and the total duration of the clonus. Manual analysis is laborious and timeconsuming, particularly for 24 hour records. Thus, the main aims of this study were todevelop an algorithm that automatically: 1) identifies when the bursts of EMG occur duringclonus, and 2) marks the start and end of these contractions. The timing of these events canbe used to calculate the duration, frequency and intensity of the contractions during theentire spasm. To be effective, this algorithm must be capable of accurately identifying theoccurrence of the repeated bursts of EMG during clonus, while at the same ignoring anymotor unit activity between these contractions. The start and end of each contraction must bedetermined irrespective of the size (magnitude or duration) of the contractions or the overallduration of the clonus. While the start and end of EMG are often measured manually inlaboratory-based studies, the number of contractions is usually limited. Isometric or constantvelocity contractions are often performed. These constraints make manual analysis feasible.In contrast, we are monitoring clonus as it occurs naturally during the day-to-day activitiesof people with SCI who have no voluntary control of leg muscles. Thus, our algorithm needsto be flexible enough to analyze data gathered under a variety of conditions, from differentmuscles, and individuals.

Only a few algorithms have been developed to identify the onset of EMG. Data have beenscanned using single or double windows, a technique that does not suppress intra-burstmaxima (Di Fabio, 1987; Marple-Horvat and Gilbey, 1992). Polynomial regression isanother approach but this requires calculation of steady state parameters and equations foreach EMG pattern (Takada and Yashiro, 1995). EMG onset has also been defined as thetime when a specified number of sample points exceed an established baseline (Di Fabio,1987; Hodges and Bui, 1996). Statistical approaches have used dual thresholds to detectmuscle activation (Bonato et al., 1998) or a generalized likelihood ratio to estimate EMGonset (Micera et al., 1998). Recent approaches have included use of wavelet transforms toeither identify motor unit action potentials at contraction onset (Merlo et al., 2003) or todetect the sudden changes in EMG that occur at the start and end of a contraction (Vannozziet al., 2010).

The algorithm developed in this study to automate the analysis of EMG during clonus alsoused wavelets for event-oriented analysis. Wavelets were scaled non-linearly to create abank of filters which were used to extract time-frequency information from the EMG (vonTscharner, 2000). To aid identification of the bursts of EMG during clonus, the algorithmwas constrained using criteria we developed from our knowledge of clonus. Signal energyrules were then applied to the globally processed EMG signals to extract the duration ofeach burst of EMG, clonus frequency and intensity. To evaluate the reliability of thealgorithm, its outputs were compared to the results produced independently by two expertswith years of experience in EMG recognition. The total amount of time taken by each

Mummidisetty et al. Page 2

J Neurosci Methods. Author manuscript; available in PMC 2013 February 15.

NIH


NIH


NIH


person to manually mark the start and end of each burst of EMG during clonus was alsocompared to the time taken by the algorithm to perform the same task. An accuratealgorithm that can automatically analyze EMG during clonus in long records is important tounderstand the nature of these involuntary muscle contractions, and the prevalence ofclonus.

2. Materials and Methods2.1 Subjects

Seven subjects (5 male, 2 female, median age: 36 yr, range: 27–52 yr) with a chronic (> 1yr) cervical SCI were studied (median time since injury: 14 yr, range: 4–33 yr). Theseinjuries were caused by diving mishaps (n=4), motor vehicle accidents (n=2), or a sportsevent (n=1). The injuries were at C4 (n=1), C6 (n=5) or C7 (n=1) and were complete (AISA) according to American Spinal Injury Association criteria (Maynard et al., 1997). Thesubjects had no voluntary control of any leg muscles, indicated by an inability to generateany voluntary EMG. Subjects took no medication to mitigate muscle spasms. All of theprocedures were approved by the Institutional Review Board of the University of Miami. Allsubjects gave informed, written consent before participating in this study.

2.2 MusclesSurface EMG signals were recorded simultaneously from 8 leg muscles over 24 hours:vastus lateralis (VL), biceps femoris (BF), tibialis anterior (TA) and medial gastrocnemius(MG) bilaterally using a bipolar configuration (Klein et al., 2010). Three self adhesiveelectrodes (Superior Silver Electrodes, Uni-patch, MN) were cut to 2.5 cm × 1.0 cm for eachmuscle. The distal electrode for TA and MG was placed just proximal to the respectivetendon-muscle interface. The distal electrode for vastus lateralis was placed 12 cm proximalto the patella. The other two electrodes for each muscle were placed proximally with aninter-electrode spacing of 4 cm. Electrodes for biceps femoris were aligned with the vastuslateralis electrodes but on the midline of the posterior surface of the leg. The electrodes weresecured to the skin with Hypafix tape (Smith & Nephew, Andover, MA) then wrapped withCo-flex bandage (Andover Healthcare, Andover, MA) to ensure that the electrodes stayed inthe same position during the entire experiment. The two distal electrodes on each muscleserved as the active and reference electrodes. The proximal electrode was the ground. Theelectrodes from each muscle were connected to the inputs of a preamplifier (Model Z03,Motion Labs Systems, Baton Rouge, LA) via soldered cables. These cables allowed customelectrode spacing, and placement of the preamplifiers on the limb close to the target muscle.Each preamplifier was wrapped in foam to provide comfortable contact with the skin over24 hours and to avoid skin breakdown.

2.3 ProtocolEach experiment consisted of a 24 hour EMG recording, and laboratory measurementsbefore and after this recording. In the laboratory, maximal compound muscle actionpotentials (M-waves) were recorded from VL, TA and MG muscles in response tosupramaximal stimulation of the femoral, common peroneal and tibial nerves, respectively.These EMG data were sampled on-line (3,000 Hz) using a SC/Zoom system (Physiologysection, Umeå University, Sweden). For each muscle, the area of the maximal M-waveswere similar before and after the 24-hour recording (Mean ± SE area for MG: 10.2 ± 1.3μVs versus 10.4 ± 1.6 μVs; TA: 6.9 ± 1.0 μVs versus 6.5 ± 0.9 μVs; VL: 9.0 ± 1.1 μVsversus 9.6 ± 1.1 μVs). These results indicate that the electrodes remained in place during theentire recording and that the changes in EMG activity over 24 hours were recorded faithfullyfrom each muscle (Klein et al., 2010). No M-waves were recorded from BF because thesciatic nerve is too deep to stimulate reliably. Subjects were also asked to perform 3 brief (5



NIH


NIH


NIH


s) maximal voluntary contractions (MVCs) with each muscle. The absence of EMG activityduring all MVCs confirmed the muscle paralysis.

The 24 hour EMG recording was initiated after the laboratory measurements using aportable, battery powered data processing and logging system (Tepavac et al., 2003). Theelectrodes from each muscle were connected to a preamplifier (Motion Labs Systems, BatonRouge, LA). The outputs from the four preamplifiers for each leg were connected to acustom-built preprocessing unit which was responsible for filtering (10–500 Hz) andamplifying the input signal to fit the input range (0 – 4.096 V; gain: ~400) of a custom-builtbattery-operated, data logging device with a12-bit analog to digital converter (Tattletale 8Logger, Onset Computer Corporation, Bourne, MA). This logger was driven by customsoftware written using Metrowerks Code Warrior (a C based software development tool;Metrowerks Corporation, Austin, TX). The sampling rate was 1000 Hz per channel. Thedata were written to a 1 GB compact flash card in compressed format.

The subject was advised to maintain his/her normal routines during the 24 hour recording toensure a representative view of daily activities, which they documented on paper. The datalogging system was stored in a hip pack that was carried on the wheelchair or the lap of thesubject. The subject returned after 24 hours to repeat the laboratory recordings, as describedabove.

2.4 Global data processingStandard data processing procedures were implemented on all 24 hour EMG recordings toenable data comparisons across muscles and experiments. This global processing occurredbefore the application of the algorithm. Using software developed in Matlab (TheMathworks Inc, Natick, MA) and DADiSP (DSP Development Corporation, Newton, MA),data from each channel were: 1) extracted to 24, 1-hour files, a manageable length forviewing; 2) set to the 24 hour clock (midnight to 1 am was designated as hour 1); 3) digitallyfiltered to eliminate noise using cascaded 30 Hz high-pass and 60 Hz FIR filters (basicwindowed linear-phase Finite Impulse Response digital filters) of order 472; 4) calibrated tomatch data gathered in the laboratory using 1 mV sine waves; 5) large artefacts, usuallyfrom cell phones or uncontrolled movements, were manually identified and excluded fromanalysis; 6) spasms that involved clonus were manually identified in the 24 hour records byan expert in EMG recognition (at least 3 contractions per spasm, each contraction 40–90 msin duration, and at a frequency of 3–10 Hz, Wallace et al., 2005).

2.5 Algorithm to automatically analyze EMG during clonusA frequency specific methodology that involved wavelets was developed here to analyze thebursts of EMG during selected examples of clonus (n=31). Only the EMG from thesespasms was subjected to various filters, as described by von Tscharner (2000), to produceintensity envelopes of the signals that both smoothed and preserved the EMG shape. TheEMG belonging to each clonus cycle (contraction) then had to be identified. Even with goodselectivity, detection of the maxima in the EMG intensity traces may produce false positivepeaks. Here, these additional peaks were effectively pruned by using rules involvingamplitude thresholds and timing constraints. Once the contraction maximum was detected,the onset and offset of each contraction was computed from the globally processed EMGsignals using signal energy rules. Thus, most frequency components of the EMG signalcontributed to the physiological characterization of clonus.

The algorithm developed here to automate the identification of contractions during clonus,as well as the start and end of the EMG for each contraction (Fig. 1A) involved 5 steps. Thisalgorithm is explained in detail in Figure 2 and the following sections:



NIH


NIH


NIH


2.5.1. Filter creation—A mother wavelet (Morlet wavelet) was scaled non-linearly toproduce a set of 10 filters with different central frequencies and frequency ranges (Table 1),as described by von Tscharner (2000).

2.5.2. Filter EMG—The EMG from 31 spasms that involved clonus was passed througheach of the filters to produce 10 intensity envelopes for a given input. To improve theprocessing time of the algorithm, the intensity envelopes were added together to obtain 3envelopes: one for the lower, intermediate, and higher frequency bands (Fig. 1B). Althoughall three envelopes followed the EMG signal, the lower frequency envelope (11.5–75.8 Hz)reacted slowly to the changes in the EMG so peaks in the EMG and the envelope weremisaligned. The higher frequency envelope (191.8–431.7 Hz) was highly reactive to thechanges in the signals but this produced many peaks in the envelope. The intermediatefrequency envelope (74.8–193.9 Hz) offered a compromise between the two other envelopesin terms of sensitivity to changes in the EMG and the production of multiple peaks.Moreover, the largest peak in the envelope for a given contraction usually occurred near themiddle of each contraction and coincided with the peak EMG amplitude. This featurefacilitated accurate determination of the time of the EMG bursts so the intermediatefrequency envelope was used for subsequent analysis (Fig. 1B).

2.5.3. Detect bursts of EMG during clonus—The crucial step of the algorithm was todetect each burst of EMG as this process accurately identified the number of contractionsthat occurred during clonus. The times at which peaks occurred in the envelope weredetermined by calculating the first derivative of the envelope and observing the changes inthe value of the slope. To ensure that each contraction was represented by only one peak, notmultiple peaks, two constraints were imposed on the algorithm:

i. An intensity threshold was established to eliminate lower peaks due to small EMGpotentials or baseline noise. For 95 % of the data, a suitable threshold was 25 μV2,determined by viewing the rectified EMG and the intensity envelope in DADiSPsoftware. The peaks in the envelope that remained after application of thisconstraint were largely confined to the burst of EMG.

ii. A 90 ms time constraint between adjacent peaks was then applied. If multiple peakswere encountered within 90 ms, the algorithm kept the peak with the maximumamplitude, which often occurred at the peak EMG (Fig. 1B). To determine this timeconstraint, the number of contractions in the various spasms involving clonus wascounted manually and compared to the peaks identified by the algorithm. The timebetween peaks was progressively changed from 40 ms to 120 ms in steps of 10 msuntil the peaks detected by the algorithm matched the manual count, whichoccurred when the separation between peaks was 90 ms.

2.5.4. Mark EMG start and end for contractions during clonus—A window wasplaced around each peak in the envelope detected by the algorithm, from the midpoint of theprevious peak to the midpoint of the next peak (Fig. 1C). The entire record was thereforecovered with consecutive windows. The total energy of the globally processed EMG signalin each window was computed using the equation:

where: X(t) is the globally processed EMG signal, Ws is the start of the window, and We isthe end of the window. When the total energy of a window was <10 mV2, a typical value fora single motor unit potential, the window was discarded, further reducing the number of



NIH


NIH


NIH


bursts of EMG identified by the algorithm. The points at which the energy in each windowreached 5 % and 95 % of the total energy were marked as EMG start and end, respectively.Energy values outside of the 5–95 % range were usually baseline noise. However in certainrecordings, motor unit potentials were present 60–80 ms before EMG onset (Fig. 1C), andthey contributed to the energy in the window. To eliminate this motor unit activity from theenergy calculation, each window (irrespective of whether a motor unit potential was presentbefore the main EMG burst) was resized to 50 ms either side of the envelope peak. The burstof EMG remained within the window because it typically lasted 40–70 ms (Dimitrijevic etal., 1980; Wallace et al., 2005). The energy contained within the resized window wasrecalculated and the starts and ends of all of the EMG bursts remarked at the revised 5–95 %values (Fig. 1C).

2.5.5. Calculation of clonus parameters—The start and end times for each burst ofEMG (contraction) during clonus were used to calculate: 1) EMG duration, the time fromthe start to the end of each burst of EMG; 2) clonus frequency (the reciprocal of the timebetween consecutive EMG starts); 3) EMG intensity, the root mean square value (RMS) ofthe EMG for each contraction, from EMG start to end; 4) clonus duration, the time from thefirst EMG start to the last EMG end during clonus.

2.6 Assessment of algorithm performanceSpasms involving clonus (n=31, 5 spasms/subject; clonus only occurred once during the 24hour recording from one person) were analyzed automatically by the program and manuallyand independently by two experts. The assessment database accounted for up to 6 % of thetotal number of spasms involving clonus during these 24 hour recordings. Data came frommedial gastrocnemius (n=20 spasms), tibialis anterior (n=4), biceps femoris (n=2), andvastus lateralis (n=5), reflecting the typical prevalence of clonus in these muscles (Wallaceet al., 2011). Results from one expert (Person 1, P1) were used as a gold standard. Algorithmperformance was evaluated by comparing the results of two people (P1; Person 2, P2) to thedata obtained from Person 1 versus the Program (Pr). Comparisons made were: 1) thenumber of common contractions (bursts of EMG) identified during each clonus; 2)agreement for EMG duration (the amount of common time measured for each contraction,calculated from each EMG start and end); 3) agreement for clonus frequency. Data for P2and the Pr were calculated as a percentage of the values obtained by P1; 4) agreement on theRMS values of the EMG, calculated for P2 and the Pr as a percentage of the RMS valueobtained by P1; 5) the time taken to measure the data. These same data were re-evaluated bymuscle to examine whether the results depended on the source of the EMG signals.

To assess whether the algorithm could be used to analyze novel data, clonus in medialgastrocnemius was analyzed in one 24 hour EMG recording from Subject F. A total of 71spasms that involved clonus were identified in this record. One of these spasms was used totrain the algorithm. The other 70 spasms were analyzed using the final algorithm. Thenumber of common contractions in these 70 new spasms, and agreement for EMG duration,clonus frequency and RMS EMG were compared for Person 1 and the algorithm, asdescribed above.

2.7 StatisticsData are presented as median (range) unless stated otherwise. Kruskal-Wallis one wayANOVA on ranks was used to evaluate agreement on the number of common contractionsmeasured with and without algorithm constraint, EMG duration, clonus frequency,contraction intensity, as well as whether the algorithm was faster than manual analysis.Intra-class correlations (ICC) were calculated to verify the inter-rater reliability of EMGdurations, clonus frequency, and RMS EMG (model-3 i.e., the two-way mixed effect



NIH


NIH


NIH


module). ICC coefficients can range from 0 (no agreement) to 1 (complete agreement) witha value greater than 0.80 regarded as satisfactory (Nunnally and Bernstein 1994). Chi-squaretests were performed to verify whether agreement between two people versus a person (P1)and the algorithm were different statistically. Statistical significance was set at p<0.05.Statistical analyses were performed using SPSS-17.0 (SPSS Inc, Chicago, IL) and Sigmastat(Systat Software, San Jose, CA).

3. Results3.1. Number of contractions during clonus

Fig. 1A shows an example of the rhythmic repeated contractions of clonus, with the EMGstart and end marked for three contractions. Person 1 manually marked the starts and ends of825 bursts of EMG during clonus (7 subjects, 31 spasms). Across subjects the mediannumber of contractions was 110 (range: 7–227; Fig. 3A). Without any constraints, thealgorithm identified a total of 4980 EMG bursts (subject median: 708, range: 48–1304),which represented a median of 686 % more contractions than Person 1 (range: 409–901 %,n=7 subjects, Fig. 3B). Two sets of constraints were applied to improve the ability of thealgorithm to detect contractions during clonus: 1) an intensity threshold to eliminate lowpeaks due to small EMG potentials or baseline noise; 2) a time constraint to prevent motorunit potentials between contractions being marked as bursts of EMG. Introduction of theintensity constraint resulted in the algorithm marking a median of 136 bursts of EMG acrosssubjects (range: 10–247 contractions, n=7). This first set of constraints reduced thecontractions marked by the algorithm to a median of 118 % of that marked by Person 1(range: 106–143 %, n=7). The second set of constraints (resizing the window placed aroundEMG bursts to eliminate motor unit potentials near the start of the contraction and removalof windows with energy as small as a single motor unit potential) bought the number ofbursts of EMG identified by the algorithm to a median of 109 across subjects (range: 8–218contractions, n=7; median: 99 %, range: 78–114 % of Person 1). Constraint of the algorithmreduced the false detection of contractions significantly (p ≤ 0.001).

3.2. Number of common contractions measured during clonusOverall Person 1, Person 2 and the constrained Program measured a total of 825, 821, and830 contractions during clonus, respectively. The median number measured across 7subjects was 112 for Person 2 (range: 7–219). Not all contractions identified by each personand the program were the same. The median agreement for common contractions measuredper subject was 99 % for two people (range: 94–100 %), and 98 % for Person 1 versus theProgram (range: 92–100 %, Fig. 3C). The agreement between two people versus Person 1and the Program did not differ significantly (p=0.946). Thus the algorithm was as effectiveas two people at identifying contractions during clonus. Differences in the markedcontractions arose because Person 1 identified low amplitude bursts of EMG which wereunmarked by both Person 2 and by the Program. Both people ignored isolated motor unitpotentials whereas the constrained program sometimes marked these potentials as bursts ofEMG.

3.3. Clonus duration and frequencyFor the 31 spasms analyzed, median clonus duration (entire spasm duration) measured by P1was 4.3 s (range: 1.1–43.4 s). The median cycle durations measured by Person 1, Person 2and the Program were 167 ms (119–208 ms, n=7 subjects), 163 ms (120–204 ms) and 172ms (119–204 ms), respectively, resulting in median clonus frequencies of 6.0 Hz (4.8–8.4Hz), 6.1 Hz (4.9–8.3 Hz), and 5.8 Hz (4.9–8.4 Hz) respectively. The median agreement forclonus frequency was 99.9 % (range: 99.0–100 %, n=7; ICC: 0.971) between Person 1 andPerson 2, and 99.7 % (range: 98.5–100 %; ICC: 0.949) between Person 1 and the Program



NIH


NIH


NIH


(Fig. 4A). The differences in clonus frequency for two people versus Person 1 and theProgram were not different statistically (p=0.718). Thus, the algorithm determined clonusfrequency as reliably as two people.

3.4. Duration of EMG during clonusThe median EMG durations (time from EMG start to end, Fig. 1A) measured by Person 1,Person 2 and the Program were 51 ms (25–56 ms), 43 ms (29–50 ms), and 43 ms (27–52ms), respectively. The median time differences were close to the average duration of surfacerecorded motor unit potentials (Thomas et al., 2006). The median common EMG timemeasured by Person 1 and Person 2 was 83.7 % (77.7–98.1 %, n=7 subjects; ICC: 0.905;Fig. 4B). Person 1 and the Program agreed 75.6 % of the time (range: 70.7–86.3 %; ICC:0.852). The differences in EMG duration measurements made by two people compared toPerson 1 versus the Program were not significant (p=0.277), indicating that the algorithmwas as good as two people at measuring the duration of EMG during clonus.

3.5. Intensity of contractions during clonusThe median RMS values for the EMG measured by Person 1, Person 2 and the Programwere 282 μV (177–404 μV, n=7), 289 μV (168–482 μV), and 288 μV (172–462 μV),respectively. The median agreement for RMS EMG was 98.3 % for Person 1 and Person 2(range: 96.9–99.6 %, ICC: 0.997) and 96.2 % for Person 1 and the Program (range: 94.9–100.0 %, ICC: 0.997; Fig. 4C). As the intraclass coefficients were the same, the Programwas as reliable at measuring contraction intensity during clonus as two people.

3.6. Analysis by muscleSimilar agreement between two people versus a person and the program was obtained whenthese 31 spasms were analyzed by muscle. Agreement for medial gastrocnemius clonusfrequency did not differ (p=0.535) between Person 1 and 2 (96.1 %) versus Person 1 and theProgram (93.0 %). Agreement for EMG duration (p=0.541; P1 versus P2: 88.4 %; P1 versusPr: 83.5 %) and RMS EMG (P1 versus P2: 99.9 %; P1 versus Pr: 99.8 %) also did not differ.Similar results were obtained for the data from the other 3 muscles (tibialis anterior, bicepsfemoris and vastus lateralis). Across these 3 muscles, clonus frequency agreement rangedfrom 85.2–98.4 % for P1 versus P2 and from 78.2–92.7 % for P1 versus Pr. Thecorresponding ranges for EMG duration agreement were 82.7–93.0 % (P1 versus P2) and72.0–90.1 % (P1 versus Pr). For RMS EMG, agreement ranged from 98.1–99.8 % (P1versus P2) and 98.0–99.8 % (P1 versus Pr). The algorithm therefore performed similarlyirrespective of the source of the EMG signals.

3.7. Measurement timeTo view and mark the start and the end of the EMG for contractions in 31 spasms involvingclonus took 5.7 hours for Person 1, 12.6 hours for Person 2, and 0.019 hours for thealgorithm. The algorithm was significantly faster at analysis than both of the humanoperators (p ≤ 0.001), who took a similar time to measure the clonus (p ≥ 0.05). As anexample, it took Person 1 1440 s (24 min) to measure the start and end of EMG in clonusthat lasted 10 s. Person 2 took 3180 s (53 min) to measure the same data, while the Programcompleted the measures in 1.64 s. These results indicate the efficiency of this novelalgorithm.

3.8. Application of the algorithm to new dataFor 70 novel spasms, Person 1 and the algorithm identified 1995 and 2035 commoncontractions during clonus, respectively, with agreement (98 %) the same as for the 31 testspasms. The median cycle durations were 165 ms (range: 112–295 ms) and 161 ms (117–



NIH


NIH


NIH


295 ms), corresponding to median clonus frequencies of 6.3 Hz (4.0–9.2 Hz) and 6.5 Hz(3.6–8.9 Hz), respectively. Agreement for clonus frequency was 99.6 %, the same as for the31 test spasms. The median EMG durations measured by Person 1 and the algorithm were44 ms (34–80 ms) and 44 ms (33–86 ms), with 96.7 % agreement for the common timemeasured. This agreement exceeded that for the test spasms. The median RMS values forthe EMG measured by Person 1 and the algorithm were 131 μV (15–395 μV) and 129 μV(15–401 μV), respectively (98.5 % agreement, which exceeded the value of 97.4 % for testspasms). Thus, for each measured parameter, median agreement for 70 novel spasms eitherequaled or exceeded that for the 31 spasms used to test the algorithm. The algorithmdeveloped here could therefore be used to accurately determine the characteristics of clonuswhen applied to novel data. The major difference was algorithm efficiency. It took 8minutes for the algorithm to analyze the clonus in 70 spasms. An additional 4.3 hours wasspent by Person 1 manually viewing, verifying and/or altering the algorithm output for everyburst of EMG during these spasms. We estimate it would take 18 hours for Person 1 to markthe start and end of every contraction during clonus in the same 70 spasms manually.

4. DiscussionThe algorithm developed in this study was able to automatically determine the location ofmuscle contractions during clonus, the duration and the intensity of the EMG for eachcontraction, and clonus frequency. The algorithm was as accurate as two people atmeasuring these parameters but performed these operations significantly faster than eitherperson. The agreement between a person and the program was highest for clonus frequency,followed by EMG intensity and duration. Efficient detection of the contractions was refinedby imposing threshold and time constraints on the algorithm. These constraints were key toaccurate characterization of clonus.

4.1 Detection of contractions during clonusThe algorithm developed here used intensity analysis to envelope the EMG in differentfrequency bands (von Tscharner, 2000). An intermediate band of frequencies (75–194 Hz),which included many of the frequency components found in surface EMG, gave an envelopethat closely followed the EMG. Peaks in the envelope typically coincided with peaks in theEMG which arose from different motor units firing asynchronously (Wallace et al., 2005).But the algorithm needed to identify only one envelope peak for each contraction so thateach EMG start and end time could be measured. This was achieved by imposing constraintsthat were developed from the known characteristics of clonus. For example, clonus involvesrhythmic contractions that repeat at a rate largely determined by the length of the reflex arc(3–10 Hz; Dimitrijevic et al., 1980; Iansek, 1984; Wallace et al., 2011). Moreover, the burstsof EMG during clonus are of relatively fixed duration, consistent with microneurographicrecordings that show muscle spindle afferents only discharge late in the muscle relaxationphase of sustained clonus (Hagbarth et al., 1975; Szumski et al., 1974). To confine theenvelope peaks to the bursts of EMG, small peaks between the contractions were discardedeffectively with the same intensity threshold. To extract a single peak per contraction, theenvelope peaks had to be separated by 90 ms (clonus frequency < 10 Hz). If there weremultiple peaks within 90 ms, the maximal peak was selected to represent a burst of EMG. Ingeneral, the constrained algorithm marked the same contractions as Person 1 (98 %agreement, range: 92–100 %). Two people agreed on 99 % (range: 94–100 %) of thecontractions. The measured data were from 7 different experiments and at least 2 differentmuscles per recording (except for one subject who experienced clonus once in 24 hours).Thus, the algorithm performed robustly across subjects and muscles despite the uncontrolledenvironment in which the data were gathered. In terms of discrepancies, both people ignoredisolated motor unit potentials, a typical way for clonus to start and end (Wallace et al.,



NIH


NIH


NIH


2005). The Program sometimes marked these potentials as bursts of EMG because theyoccurred at clonus frequency, adhered to the 90 ms time constraint, and had sufficientenergy to be retained. Second, Person 1 identified low amplitude bursts of EMG, also typicalat the start and end of clonus, which were unmarked by the Program or Person 2. In somecases these low bursts of EMG fell below the intensity threshold set for the program.

4.2 Determination of contraction frequency, duration and intensity during clonusKnown characteristics of clonus were also used to guide how the algorithm determined thestart and end of EMG for each contraction during clonus. The EMG in each contractiontypically rises then declines in amplitude over 40–70 ms (Dimitrijevic et al., 1980), withmost motor units firing once during this contraction. In some cases, single motor unitresponses were 60–80 ms before the contraction (Wallace et al., 2005). The algorithmcaptured the main contraction and avoided early motor unit potentials by centering windowsaround the peaks in the intensity envelopes (± 50 ms). This window just exceeded thetypical duration of EMG. The window was usually placed symmetrically on the contractionbecause the envelope and EMG peaks aligned closely with the maximal peak near the centerof the contraction. The times where the energy contained within the resized window reached5 % and 95 % of the maximal value were optimal for the EMG start and end, respectively.

When motor unit potentials fell within the resized window but just outside of the maincontraction they were included in the energy calculation used to determine start and endtimes. The differences in EMG durations measured by Person 1 and the Program or by twopeople were 8 ms, close to the average duration of motor unit potentials measured fromsurface EMG (Thomas et al., 2006). In contrast, a potential with energy smaller than 5 % ofthe total energy will be excluded. Thus, the chance of inappropriate inclusion of a motor unitpotential is higher when the intensity of the EMG burst is low. Besides inclusion of isolatedmotor unit potentials in the main contraction, some contractions included large amplitudepotentials, possibly due to motor units responding in synchrony. In these cases the Programstarted and ended the EMG early because the majority of the energy was contributed by thebig potential alone. This resulted in short EMG durations. In addition, the filters used toeliminate noise and artefacts from the entire 24 hour recording produced a slow wave at thestart and end of some contractions and these waves were included by Person 1. The Programexcluded these waves because they were less than 5 % of the energy of the window.

Despite these differences, the algorithm was as accurate as two people at determining clonusfrequency, duration and intensity. This was the case when data were analyzed by subject orby muscle. Agreement was probably highest for clonus frequency because it was calculatedonly from the start of the EMG whereas EMG duration involved both the EMG start and endtimes. Signals smaller than 5 % of maximal energy were largely small EMG potentials,baseline noise, or slow waves from filters. As a consequence, small differences in EMGduration measurements had minimal effect on assessments of contraction intensity measuredby people and the algorithm across spasms, muscles and experiments.

Other methods that use wavelet transform techniques to detect the start and end of EMGreport accuracy similar to the current study. When contraction start was estimated byidentifying motor unit potentials in simulated EMG signals, the standard deviation of theonset estimate was 17 ms for signal to noise ratios between 2 and 8 dB (Merlo et al., 2003).Detection of EMG onset and offset varied between 7 and 11 ms, respectively, when assessedfrom sudden changes in the EMG signal (Vannozzi et al., 2010). With our method, onset andoffset varied between 9 ms and 12 ms, respectively. Implementation of the methoddeveloped by Vannozzi et al. (2010) is more complex computationally (multiple modules)than our approach and requires an efficient reasoning module to choose accurate onset andoffset times. It would be informative to use this approach in future studies as it may provide



NIH


NIH


NIH


a slightly more accurate estimate of EMG onset and offset. But only an incrementalimprovement in the measurement of clonus frequency can be expected. Current agreementbetween the program and an EMG expert ranged from 98–100 %. Nevertheless, the fast,rhythmic contractions of clonus are likely to be an important feature that disrupts dailyactivities.

4. 3 Analysis timeThe algorithm was 574 times faster than human operators at locating and measuring the startand end times of bursts of EMG during clonus. An algorithm that is accurate and fast isinvaluable for processing large volumes of data. Even though a person needs to verify theoutput produced by the algorithm when analyzing 24 hour records, the time needed tocorrect outliers is still 4–5 times less than manual analysis alone. Moreover, humanoperators measure optimally when the clonus duration is short. A person has to be consistentin their measurements and decision making to measure clonus of long duration (>5 s). Thistask requires concentration. Otherwise, mistakes are made. Completing this analysis on aregular basis is also laborious and uninteresting. Thus, it is ideal to automate such laborintensive processes to ensure accurate data analysis.

4.4 Analysis of novel dataThe final algorithm developed here was used successfully to analyze 70 new examples ofclonus with high accuracy. Agreement between Person 1 and the Program either equalled orexceeded that obtained for the 31 test spasms for identification of common contractions,clonus frequency, EMG duration and intensity. Furthermore, the algorithm completed theanalysis in minutes. Verification of the results took 4 hours. Visually scanning the EMGrecords took most of this time. Result correction took little time. In contrast, our experiencewith manual analysis suggests it would take days for a person to perform this analysismanually. They would then need to verify their results. Thus, this novel algorithm is avaluable resource for timely, accurate analysis of clonus in long duration EMG records.

4.5 Limitations and future developmentsThere are no existing methods to analyze clonus automatically in long-term (24 hour) EMGrecordings. The algorithm developed in this study is one possible approach to this laborintensive task. It was able to successfully process and analyze test and novel spasms inseconds. It worked best when there were no motor unit potentials between contractions. Thissituation arose in 5 of the 7 recordings. Algorithm constraint helped eliminate falseidentification of motor unit potentials as contractions. A user could also easily eliminatethese extra contractions. Future research could enhance automation further. For example,intensity thresholds to discard small amplitude signals could be adapted depending on thecomplexity of the EMG. When motor unit potentials were prevalent between thecontractions, we found peaks in the envelopes could be reduced by switching to the lowfrequency envelope. Use of this alternative envelope helped to locate the contractions ofinterest. Automatic identification of where clonus occurs in 24 hour records could also beimplemented.

4. 5 Functional implicationsThe number of bursts of EMG during clonus is of clinical importance because it conveysinformation about clonus duration. A longer spasm may make it more difficult to performdaily activities (Adams and Hicks, 2005, Little et al., 1989, Sheean, 2002). RMS EMG canbe used to estimate the strength of contractions during clonus, an unexplored area. Strongcontractions may be more disruptive to the person. Clonus can occur in a single muscle orsimultaneously in multiple muscles (Cook 1967). Analyzing clonus in multiple muscles may



NIH


NIH


NIH


allow the resultant limb movement to be predicted. Furthermore, analyzing clonus in long-term (24 hr) EMG recordings may provide a valuable way to evaluate interventions designedto mitigate clonus.

AcknowledgmentsThe authors thank Sean Ferrell and Dr. Lourdes Silva for identifying clonus in 24 hour EMG records, AdrianaMartinez for help with the references and Yenisel Cruz-Almeida for statistical help. This research was funded byNational Institutes of Health grant NS30226, The Miami Project to Cure Paralysis, and the Department ofBiomedical Engineering, University of Miami.

ReferencesAdams MM, Hicks AL. Spasticity after spinal cord injury. Spinal Cord. 2005; 43:577–86. [PubMed:

15838527]Beres-Jones JA, Johnson TD, Harkema SJ. Clonus after human spinal cord injury cannot be attributed

solely to recurrent muscle-tendon stretch. Exp Brain Res. 2003; 149:222–36. [PubMed: 12610691]Bonato P, D’Alessio T, Knaflitz M. A statistical method for the measurement of muscle activation

intervals from surface myoelectric signal during gait. IEEE Trans Biomed Eng. 1998; 45:287–99.[PubMed: 9509745]

Cook WA Jr. Antagonistic muscles in the production of clonus in man. Neurology. 1967; 17:779–96.[PubMed: 15088538]

Di Fabio RP. Reliability of computerized surface electromyography for determining the onset ofmuscle activity. Phys Ther. 1987; 67:43–8. [PubMed: 3797476]

Dimitrijevic MR, Nathan PW, Sherwood AM. Clonus: the role of central mechanisms. J NeurolNeurosurg Psychiatry. 1980; 43:321–32. [PubMed: 6445412]

Hagbarth KE, Wallin G, Löfstedt L, Aquilonius SM. Muscle spindle activity in alternating tremor ofParkinsonism and in clonus. J Neurol Neurosurg Psychiatry. 1975; 38:636–41. [PubMed: 125783]

Hodges PW, Bui BH. A comparison of computer-based methods for the determination of onset ofmuscle contraction using electromyography. Electroencephalogr Clin Neurophysiol. 1996;101:511–9. [PubMed: 9020824]

Iansek R. The effects of reflex path length on clonus frequency in spastic muscles. J Neurol NeurosurgPsychiatry. 1984; 47:1122–4. [PubMed: 6094731]

Klein CS, Peterson LB, Ferrell S, Thomas CK. Sensitivity of 24-h EMG duration and intensity in thehuman vastus lateralis muscle to threshold changes. J Appl Physiol. 2010; 108:655–61. [PubMed:20007856]

Little JW, Micklesen P, Umlauf R, Britell C. Lower extremity manifestations of spasticity in chronicspinal cord injury. Am J Phys Med Rehabil. 1989; 68:32–6. [PubMed: 2917056]

Marple-Horvat DE, Gilbey SL. A method for automatic identification of periods of muscular activityfrom EMG recordings. J Neurosci Methods. 1992; 42:163–7. [PubMed: 1501501]

Maynard FM Jr, Bracken MB, Creasey G, Ditunno JF Jr, Donovan WH, Ducker TB, Garber SL,Marino RJ, Stover SL, Tator CH, Waters RL, Wilberger JE, Young W. International Standards forNeurological and Functional Classification of Spinal Cord Injury. American Spinal InjuryAssociation Spinal Cord. 1997; 35:266–74.

Merlo A, Farina D, Merletti R. A fast and reliable technique for muscle activity detection from surfaceEMG signals. IEEE Trans Biomed Eng. 2003; 50:316–23. [PubMed: 12669988]

Micera S, Sabatini AM, Dario P. An algorithm for detecting the onset of muscle contraction by EMGsignal processing. Med Eng Phys. 1998; 20:211–5. [PubMed: 9690491]

Nunnally, JC.; Bernstein, IH. Psychometric theory. 3. McGraw-Hill; New York: 1994. p. 83-113.Rack PM, Ross HF, Thilmann AF. The ankle stretch reflexes in normal and spastic subjects. The

response to sinusoidal movement Brain. 1984; 107:637–54.Rossi A, Mazzocchio R, Scarpini C. Clonus in man: a rhythmic oscillation maintained by a reflex

mechanism. Electroencephalogr Clin Neurophysiol. 1990; 75:56–63. [PubMed: 1688774]Sheean G. The pathophysiology of spasticity. Eur J Neurol. 2002; 9:S3–S9.



NIH


NIH


NIH


Szumski AJ, Burg D, Struppler A, Velho F. Activity of muscle spindles during muscle twitch andclonus in normal and spastic human subjects. Electroencephalogr Clin Neurophysiol. 1974;37:589–97. [PubMed: 4140072]

Takada K, Yashiro K. Automatic measurement of on/off periods of EMG activity. Medinfo. 1995;8:751–4. [PubMed: 8591318]

Tepavac, D.; Thomas, CK.; Winslow, J.; Klein, CS.; Gan, X.; Dikshit, A. Long-term surface EMGrecording system. Biomedical Engineering Society meeting; 2003. p. 24

Thomas CK, Johansson RS, Bigland-Ritchie B. EMG changes in human thenar motor units with forcepotentiation and fatigue. J Neurophysiol. 2006; 95:1518–26. [PubMed: 16267110]

Vannozzi G, Conforto S, D’Alessio T. Automatic detection of surface EMG activation timing using awavelet transform based method. J Electromyogr Kinesiol. 2010; 20:767–72. [PubMed:20303286]

von Tscharner V. Intensity analysis in time-frequency space of surface myoelectric signals by waveletsof specified resolution. J Electromyogr Kinesiol. 2000; 10:433–45. [PubMed: 11102846]

Wallace DM, Ross BH, Thomas CK. Motor unit behavior during clonus. J Appl Physiol. 2005;99:2166–72. [PubMed: 16099891]

Wallace DM, Ross BH, Thomas CK. Characteristics of lower extremity clonus after human cervicalspinal cord injury. J Neurotrauma. 2011 in press.

Walsh EG. Clonus: beats provoked by the application of a rhythmic force. J Neurol NeurosurgPsychiatry. 1976; 39:266–74. [PubMed: 932741]



NIH


NIH


NIH


Research Highlights

• Novel algorithm to automatically characterize clonus in long-term EMG records

• Wavelets were scaled non-linearly to extract time-frequency information fromEMG

• Threshold and time constraints improved algorithm accuracy

• Robust algorithm performance across muscles, subjects, and time

• Potential research tool to facilitate analysis of involuntary muscle contractions



NIH


NIH


NIH


Fig. 1.EMG during clonus. A. Three cycles of clonus with the start (---) and end (…) of theunrectified EMG marked for each contraction. The EMG were recorded from the left medialgastrocnemius muscle (hour 15: 3–4 pm) of a subject with a SCI at C4. B. The same EMGrectified and with an overlay of the linear envelopes generated for the high, intermediate andlow frequency bands. Arrows show the EMG peaks identified by the algorithm withoutconstraints. C. The EMG from one contraction enclosed by a default window (---), a resizedwindow (…), and the start and end of the EMG (__) determined by the algorithm for theresized window. Shortening the window excluded the motor unit potential between thecontractions.



NIH


NIH


NIH


Fig. 2.Flow chart of the steps the algorithm implemented to automatically identify contractionsduring clonus and to mark the starts and ends of the EMG. EMG (globally processed) wasinput to the algorithm twice.



NIH


NIH


NIH


Fig. 3.Algorithm performance with and without constraints. A. Median number of contractionsidentified by the algorithm with no constraints, one, or two sets of constraints (filled bars)compared to the analysis of Person 1 (n=31 spasms, n=7 subjects, 5 spasms per subjectexcept when clonus occurred once in a 24 hour recording). B. Same data as in A, expressedas a percentage of the results generated by Person 1. C. Median (25th and 75th percentiles)agreement for common contractions measured by Person 1 (P1) versus Person 2 (P2)compared to Person 1 versus the Program (Pr). Data for each subject are shown withdifferent symbols.



NIH


NIH


NIH


Fig. 4.Agreement on clonus frequency, duration and intensity. The median (25th and 75thpercentiles) agreement for clonus frequency (A), EMG duration (B), and RMS EMG (C)analyzed by Person 1 (P1) and Person 2 (P2) versus Person 1 and the Program (Pr, n=7subjects). Results for each subject are shown by unique symbols.



NIH


NIH


NIH


NIH


NIH


NIH



Table 1

Frequencies used to create filters

Wavelet index (j)

Central

Frequency (Hz)

Minimum Maximum

1 19.3 11.5 27.1

2 37.7 27.0 48.5

3 62.1 48.5 75.8

4 92.4 74.8 110.0

5 128.5 108.0 149.0

6 170.4 147.0 193.9

7 218.1 191.8 244.5

8 271.5 242.2 300.8

9 330.6 297.4 363.8

10 395.5 359.4 431.7


Date post:	10-Nov-2023
Category:	Documents
Upload:	independent
View:	0 times
Download:	0 times

Automatic analysis of EMG during clonus

Documents