Advanced Biomechanics of Physical Activity KIN 831 Electromyography

Advanced Biomechanics of Physical Activity (KIN 831) Electromyography (EMG) Material included in this presentation is derived primarily from two sources: * http: //www. delsys. com/library/tutorials. htm * Nigg, B. M. & Herzog, W. (1994). Biomechanics of the musculo-skeletal system. New York: Wiley & Sons * Winter, D. A. (1990). Biomechanical and motor control of human movement. (2 nd ed. ). New York: Wiley & Sons

Electromyography (EMG) • Electro – electrical • Myo – muscle • Graphy – record ----------------------------Electromyography – involves recording the electrical activity of muscle Electromyogram – electrical signal associated with the contraction of a muscle

Selected Historical Events Related to EMG • Andreas Vesalius, “father of modern anatomy”, appearance and geography of dead muscle, 1555

Selected Historical Events Related to EMG • William Croone, in De Ratione Motus Musculorum concluded from nerve section experiments that the brain must send a signal to the muscles to cause contraction, 1664 • Physiologists became excited over the phenomena produced by electrical stimulation of muscles, 1740

Selected Historical Events Related to EMG • Albrecht von Haller (1708 -1777) summarized many of the earlier studies in his treatise on muscular irritability. • Robert Whyatt (1714 -1766) reported clinical observations on a patient treated by electrotherapy. • “Animal electricity” was proposed as a substitute for the “animal spirits” which earlier experiments believed to be the activating force in muscular movement.

Selected Historical Events Related to EMG • Luigi Galvani (1737 -1798) studied the effects of atmospheric electricity upon dissected frog muscles. He concluded that the movement of the muscle was the result of its exterior negative charge uniting with the positive electricity which proceeded along the nerve (1786). Galvani’s Commentary on the Effects of Electricity on Muscular Motion (1791 or 1792) is probably the earliest statement of the presence of electrical potentials in nerve and muscle. He showed that electrical stimulation of muscular tissues produced contraction and force. He is considered the “father of experimental neurology. ”

Galvani’s demonstrations of the effects of electricity on muscles of frogs and sheep (De viribus electricitatis in motu musculari commentarius, 1792)

Selected Historical Events Related to EMG • “animal electricity” became the absorbing interest of the physiological world. The greatest name among the early students of the subject was Emil Du. Bois-Reymond (1818 -1896). He laid the foundation of modern electrophysiology. He was probably the first to discover and describe that contraction and force production of skeletal muscle were associated with electrical signals originating from the muscles (1849).

Selected Historical Events Related to EMG Guillaume Benjamin Amand Duchenne (1806 -1875) set out to classify the functions of individual muscles through electrical stimulation. He recognized the problem of attempting to isolate muscle contractions.

Duchenne’s book, Physiology des Movements (1865), has been acclaimed “one of the greatest books of all time. ” He was probably the first to perform systematic investigations of muscular function using an electrical stimulation approach.

Guillaume Benjamin Amand Duchenne de Boulogne investigating the effect of electrical stimulation of the left frontalis muscle on one of his cooperative (prisoner) subjects

Selected Historical Events Related to EMG • Wedinski (1880) demonstrated the existence of action currents in human muscle. Practical use had to await the invention of a sensitive string galvanometer (W. Einthoven - 1906).

Selected Historical Events Related to EMG • The physiological aspects of EMG were first discussed (1910 -1912) by H. Piper of Germany • E. D. Adrian, in an article in Lancet (1925, vol. 2, pp. 1229 -1233) entitled “Interpretation of the Electromyogram” demonstrated for the first time that it was possible to determine the amount of activity in a human muscle at any stage of movement.

Selected Historical Events Related to EMG • Toward the end of WWII, with marked improvement of electronic apparatus; anatomists, kinesiologists, and orthopedic surgeons began to make increasing use of EMG. The first study that gained wide acceptance was that of Inman, Saunders, and Abbott who reported their work on the movements of the shoulder region in “Observation on the Function of the Shoulder Joint” in the Journal of Bone and Joint Surgery (1944, vol. 26, pp. 1 -30).

We have come a long way!!!

Selected Historical Events Related to EMG • During the 1950’s and beyond , EMG for kinesiological studies became widespread.

EMG of normal gait? ? ? Note the use of event markers in the foot.

Selected Historical Events Related to EMG • John Basmajian (1921 - ) wrote the bible of electromyography entitled Muscles Alive. He and Carlo De Luca summarized the existing knowledge and research on muscle function as revealed by EMG studies.

Copper screen cage to inhibit noise in the EMG signal

Typical multifactorial gait-recording showing: A. Angular accelerometer on the left leg B. Vertical accelerometer C. Horizontal accelerometer D. Strain gauge tensiometer on left gastrocnemius E. EMG of left gastrocnemius

Electromyography is a seductive muse because it provides an easy access to physiological processes that cause the muscle to generate force, produce movement and accomplish the countless functions which allow us to interact with the world around us. The current state of Surface Electromyography is enigmatic. It provides many important and useful applications, but it has many limitations which must be understood, considered and eventually removed so that the discipline is more scientifically based and less reliant on the art of use. To its detriment, electromyography is too easy to use and consequently too easy to abuse. C. J. De Luca, 1993

Schematic Representation of a Recording an EMG Signal from a Single Muscle Fiber • Measure of changes in electrical potential across the muscle fiber • At rest, potential ≈ -90 mv • With sufficient stimulation potential inside cell rises to ≈ 30 -40 mv • Change in potential (fiber action potential) can be recorded • Action potentials from multiple fibers in a motor unit are simultaneously recorded • Signal from depolarization of a motor unit is called motor unit action potential

Electrophysiology of Muscle Contraction 1. Motor unit action potential (muap) – change in electrical potential across the muscle fiber membranes when a motor unit is stimulated beyond a critical threshold 2. Electrodes placed inside (indwelling) or on the surface of a muscle record the algebraic sum of all muap’s transmitted along muscle fibers that reach the electrodes 3. Motor units far away from the electrode have their muap attenuated (i. e. , are smaller) 4. Motor units of a muscle are controlled by motor neurons activating them at their motor end plates

Electrophysiology of Muscle Contraction 5. End plate potential (EPP) – depolarization of post synaptic membrane 6. EPP that reach a threshold initiate action potential in muscle fiber membrane 7. Depolarization of the transverse tubular system and sarcoplasmic reticulum results in a depolarization wavealong the direction of the muscle fibers 8. EMG records the depolarization and subsequent repolarization

Two Categories of Electrodes 1. By placement of electrode: • Surface • Indwelling (needle)

Delsys Surface Electrodes

Delsys Surface Electrodes

Comparison between Recording Areas of Two Types of Surface Electrodes

• Indwelling (needle) Steps in making a bipolar fine -wire electrode (Basmajian and Stecko, 1962)

Surface vs. Indwelling Electrodes • Surface – Non-invasive – Detect average activity of superficial muscles and give more reproducible results – Metal (silver/silver chloride) disk or bar – May be subject to cross -talk (EMG signals from motor units of other muscles near by • Indwelling – Invasive – Used to detect EMG signal from small muscles and deep muscles – Fine hypodermic needle with insulated wire conductors – May be subject to cross -talk

Preparation of Skin for Surface Electrodes 1. Reduce electrical impedance of skin • • Shave the area Apply rubbing alcohol or abrasives to remove dead skin and oils 2. Use electrode gel and pressure, adhesive tapes and/or elastic bands to affix electrode to skin

Categories of Electrodes 2. By electrode configuration: • Monopolar – records difference in voltage relative to ground • Bipolar – two contacts to measure electrical potential, each relative to a common ground, most common electrode type • Multipolar

Biphasic Signal associated with single electrode and ground

Triphasic Signal associated with voltage difference when two electrodes are used at one site

Factors Affecting EMG Signal • Propagation velocity of wave front (≈ 4 m/s) – Fatigue results in decreased propagation velocity • Distance between electrodes • Depth of muscle fibers being recorded • Electrode surface area – Larger surface area longer duration of muap – surface electrodes record longer muap than indwelling electrodes (≈ 3 -20 ms) • Size of muscle fibers being recorded – Larger fibers have larger signals

Preferred electrode location is between motor point (innervation zone) and the tendonous insertion.

Amplitude and frequency spectrum of EMG signal affected by electrode placement with respect to: A Myotendonous junction D B, C Edge of muscle B C Preferred location: D Midline of belly between innervation zone and myotendonous junction - greatest amplitude detected A

Factors to Consider in Recording EMG Signals • EMG signal is summation of muap’s • Goal is to have signals that are undistorted (linear amplification) and free of noise (biological – ECG, other muscles; man-made – power lines, machinery) and artifacts (false signals from electrodes and cabling – movement artifacts from touching electrodes or moving cables) • Large signals 5 -10 m. V; small signals 100 V

Factors to Consider in Amplifying EMG Signals • Amplifier gain – ratio of output voltage to input voltage (gain of 1000: 2 m. V 2 V) – Linear amplification over entire band width – Do not overdrive the amplifier system (large signals clipped off) – Full range frequency response for amplifier should be fast enough to handle highest EMG frequencies • Amplifier input impedance –resistance – High so as not to attenuate the EMG signal *Report magnitudes of voltage as they are sensed at the electrodes; not amplified signal

Factors to Consider in Amplifying EMG Signals • Frequency response – Amplify without attenuation all frequencies • Frequency spectrum of EMG signals from 5 to 2000 Hz • Recommended range for surface electrodes – 10 to 1000 Hz • Recommended range for indwelling electrodes – 20 to 2000 Hz – Bandwidth of amplifier difference between upper and lower cutoff frequencies – Possible filtering of signals to avoid unwanted noise

Want frequencies of EMG signals to fall within range where all frequencies are linearly influenced by gain 5 Hz 2000 Hz

Power density spectrum – mathematical conversion of EMG signals from time to frequency domain for analysis of the frequency content of the signal • Higher frequency content of indwelling electrodes because of closer spacing of electrodes and their closer proximity to active muscle fibers • Most of EMG signal concentrated in band width between 20 and 200 Hz • Problem with power lines because frequency is in middle of band width • Movement artifact (0 -10 Hz) can be filtered without adversely affecting desired EMG signal

Factors to Consider in Amplifying EMG Signals • Common mode rejection – Human body good conductor; acts as antenna to electromagnetic radiation – Want to eliminate extraneous signals – Unwanted signals picked up simultaneously at two locations can be eliminated resulting in amplification of only difference in voltage associated with EMG signal • Desired amplified signal = A[(Vhum + emg 1) - (Vhum + emg 2)] = A[emg 1 – emg 2]

Analog to Digital Conversion and Sampling an Analog Signal

Analog EMG signal Digital display of analog EMG signal sampled at 2 k. Hz

Sampling a 1 V, 1 Hz sinusoid at 10 Hz Recreating the sinusoid at 10 Hz

Sampling a 1 V, 1 Hz sinusoid at 2 Hz Recreating the sinusoid at 2 Hz

Sampling a 1 V, 1 Hz sinusoid at 4/3 Hz Recreating the sinusoid sampled at 4/3 yields a 1/3 Hz signal. The original 1 Hz signal is undersampled.

The Nyquist Frequency Signals should be sampled at no less than twice the original frequency.

Fourier decomposition of maup • Original signal in red • Superimposed signal in blue is the mathematical summation of the 10 sinusoids above • Exact reconstruction would require an infinite number of sinusoids, but 10 provides appropriate accuracy time Signal is in time domain because it expresses voltage as a function of time.

Signal of muap from previous slide is in the frequency domain because it describes amplitudes of the frequency contained in it.

Unprocessed EMG Signals • Useful for determining: – Onset and turn-off of muscle contraction – Pattern of contraction of muscles – Electromechanical delay (EMD)

Why Process EMG Signals? • Raw signals resemble noise (stochastic) • Raw signals fluctuate around 0 voltage ( V over time 0) V over time for all EMG records are the same; no differentiation • Processed signals may be correlated to parameters of muscle contraction being studied (e. g. , force, fatigue)

Processing EMG Signals in the Time Domain • Rectification – Half wave – eliminate negative values; only positive signals are used – Full wave – absolute value of all signals used • Preferred because no information is eliminated • Often used in further processing • Smoothing – Filtering signal to eliminate selected frequencies • Low pass filter – allows low frequencies to pass untenanted, but removes most of the high frequencies • High pass filter – allows high frequencies to pass untenanted, but removes most of the low frequencies • Window or notch filter

Some Common EMG Processing Absolute value of EMG signal Full wave rectified and low pass filter Area under voltage time curve with time reset

Examples of EMG Signal Processed in the Time Domain

Processing EMG Signals in the Time Domain • Integration – measures the area under the volt-time curve IEMG = • Reset at regular intervals of time • Reset at regular intervals of pre-established area (Vsec)

Processing EMG Signals in the Time Domain • Root Mean Square – Frequently used in studying muscular fatigue – Calculation • Sum of squared raw data values of EMG signal • Determine mean of sum • Take square root of the mean RMS =

Processing EMG Signals in the Frequency Domain • Power density spectra – Frequency domain important because frequency content of EMG signal shown to be reduced with fatigue – Power density spectra of EMG signal obtained using Fast Fourier Transformation technique – Mean and median frequency, bandwidth, and peak power frequency examples of use of power density spectra

Example of EMG Signal Processed in the Frequency Domain Frequency spectrum of EMG signal detected from the tibialis anterior muscle during a constant force isometric contraction at 50% voluntary maximum.

Power density spectrum of EMG signal obtained from Fast Fourier Transformation (FFT)

Mean and Median Frequencies • Mean frequency – that frequency where the product of the frequency value and the amplitude of the spectrum is equal to the average of all such products throughout the complete spectrum; used mainly to monitor muscle fatigue • Median frequency – that frequency that divides the power density spectrum into two regions having the same amount of power; preferred for detecting muscle fatigue – Less sensitive to signal noise – Less sensitive to aliasing – More often more sensitive to biochemical and physiological factors in muscle during sustained contractions

Meaning of EMG Signals • • Logical to assume that EMG signals relate to biomechanical variables (e. g. , muscle contraction force, muscle fatigue) Quandary: EMG signal is the result of many physiological, anatomical, and technical factors

Meaning of EMG Signals • 5 cardinal questions 1. Is the signal detected and recorded with maximum fidelity? 2. How should signal be analyzed? 3. Where does the detected signal originate? (cross talk, electrode placement on muscle) 4. Is signal stationary? 5. Where does the measured force originate? (influence of synergists and antagonists)

Relationships between EMG Signals and Biomechanical Variables - Force • Qualitative relationship not questioned in scientific literature; quantitative nature hotly debated • Quantitative relationship difficult to show – Difficulties measuring EMG and force of muscle contraction – Problem with temporal disassociation of muscular contraction and EMG signal (EMD)

Relationships between EMG Signals and Biomechanical Variables – Force • Isometric contraction – Can eliminate problems with measurement of force of contraction and EMG – Can eliminate temporal dissociation by sampling in middle of steady state contraction – Despite ability to eliminate or reduce problems • Different relations between force and EMG seen – – Muscle specific relationships with EMG? Force measured indirectly? Activity of antagonists or synergists? Signal processed differently in each study – Linear and non-linear relationships found

Electromechanical Delay (EMD)

Rat Muscle Soleus – slow twitch, high aerobic, slow fatiguing Extensor digitorum longus – fast twitch, high glycolytic, fast fatiguing *Note dramatic delay of force time rise under same stimulation conditions

Relationships between EMG Signals and Biomechanical Variables – Force • Dynamic contractions (concentric, eccentric, isokinetic) – Few studies with unrestrained movement – Because of problems, most studies of isokinetic contraction • Constant angular velocity constant velocity of muscle shortening • Constant angular velocity constant velocity of contractile element shortening – EMG amplitude associated with negative work considerably less than positive work

Relationships between EMG Signals and Biomechanical Variables –Fatigue • Fatigue – “point” at which force of contraction can not be maintained • Problems in measuring fatigue – Which muscle is fatigued? – Variable recruitment and utilization of motor units – Fatigue both psychological and physiological phenomena

Relationships between EMG Signals and Biomechanical Variables –Fatigue • Fatigue is associated with a shift in the frequency spectrum of the EMG signals to lower frequencies – Lower conduction velocities of some or all action potentials – Slower motor units remain active while faster motor units drop out – Motor units tend to fire more synchronously

• Diagrammatic explanation of spectral modification which occurs in EMG signal during sustained contractions • Muscle fatigue index is represented by the median frequency of the spectrum

Factors Causative Extrinsic • Electrode Intermediate • Differential electrode filter • Motor point • Detection volume • Fiber orientation • Tendon Intrinsic • Number of active motor units • Motor unit firing rate (synchronization) • Fiber type Lactic acid (p. H) • Signal crosstalk • Number of motor units detected • Muscle activation (on/off) • MUAP amplitude • Conduction velocity • MUAP shape • Fiber diameter • Subcutaneous tissue • Amplitude (RMS/ARV) • MUAP duration • Blood flow • Electrode Fiber location • Muscle fiber (net force/torque) • Muscle fiber interactions • Motor unit firing rate • Superposition Interpretation • Number of active motor units • Motor unit twitch force • Configuration • Muscle edge Deterministic EMG Signal • Spatial filtering • Recruitment stability • Spectral variables (median/mean frequency) • Muscle fatigue • Muscle biochemistry • Other factors Schematic of factors affecting EMG signal – influences and interactions, C. J. De Luca, 1993