Overview of EMG

Electromyography (EMG) is the recording of electrical activity from muscles. The basic premise of EMG is that depolarization of a motor nerve produces a recordable electrical potential within one or more muscles innervated by that particular nerve. EMG activity is recorded using needle electrodes placed subdermally (under the skin and near a muscle), or directly in the belly of the muscle(s) of interest.


EMG is recorded with electrodes placed into the muscle of interest.

EMG is recorded with electrodes placed into the muscle of interest.


Note: the use of automated or semi-automated EMG monitoring devices is dangerous and not recommended. Electrical stimulation of nervous system tissue and recording of EMG activity should be performed by a professional neurophysiologist.

Free-Running EMG Recordings:

Free-running EMG provides real-time feedback whenever a motor nerve is activated or irritated. Accurate interpretation of EMG is facilitated by simultaneous visual and auditory monitoring, so a speaker is used in parallel with a visual display to provide concurrent auditory feedback to the neurophysiologist and surgeon. While the EMG recording is continuous, a snapshot of a computer screen during a recording might look like this:

EMG signal recorded from the bicep during two brief muscle contractions.

EMG signal recorded from the bicep during two brief muscle contractions.

Neurotonic EMG activity (aka A-train EMG), characterized by irregular, high frequency (50-300 Hz) burst and train activity, is of greatest concern as it may be caused by nerve compression, traction, or blunt trauma. By contrast, relatively regular, low frequency activity is usually benign and in some circumstances may indicate insufficient patient hypnosis or reactivity to painful stimuli.

EMG trains lasting longer than 10 seconds have been associated with postoperative deficit; however, absence of spontaneous EMG activity is not necessarily indicative of intact nerve function. Indeed, neurotonic EMG may be absent following serious nerve injury, including sharp dissection. Thus, spontaneous EMG has limited sensitivity to nerve injury.

Stimulus-Triggered EMG Recordings:

Stimulus-Triggered EMG is recorded in response to direct electrical stimulation of nerves or other tissue. A hand-held probe is used to deliver electrical current to the site of interest. When a functional nerve is depolarized, a response is recorded in the form of a compound muscle action potential (CMAP). The recording window is time-locked to the onset of the stimulus, allowing the latency and amplitude of CMAPs to be quantified and compared. A snapshot of a computer screen during a recording might look like this:

CMAP recorded from a single muscle after delivery of electrical stimulation.

CMAP recorded from a single muscle after delivery of electrical stimulation. From: http://www.scielo.br/scielo.php?pid=S1808-18512013000400013&script=sci_arttext&tlng=en


In order to interpret their meaning, all EMG recordings must be quantified.  Thus, we look at the following variables:


Amplitude of EMG is an indirect measure of how much of the nerve is activated. It is measured in μV. High amplitude free-running EMG may be indicative of significant nerve irritation. Some people use stimulus-triggered EMG to evaluate nerve function by comparing the maximum achievable CMAP amplitude before and after surgery. Reduced CMAP amplitude may be indicative of injury.


The threshold is operationally defined as the minimum current necessary to trigger a CMAP with amplitude > 0 μV. Threshold is measured in mA. Some people use stimulus-triggered EMG to evaluate nerve function by comparing the CMAP threshold before and after surgery. Threshold can also be used to differentiate between sensory and motor nerves. 


When we talk about latency of EMG in surgery, we are usually referring to Stimulus-Triggered EMG recordings. How long does it take the muscle to respond after activation of the nerve? Latency is measured in msec. Latency can be used to identify certain nerves, and to differentiate between sensory and motor nerves. Conduction time is slowed by low temperatures, nerve compression, nerve injury, etc.  So, increases in latency can be indicative of evolving injury.


Frequency refers to the number of spikes that occur over the course of a second in a free-running EMG recording. Frequency is measured in Hz.

Train Time:

When EMG activity is observed in a free-running EMG recording, train time refers to the time from onset to resolution. Train time is measured in seconds or minutes.

Why we use it in surgery:

1. Monitor nerves for stretch/irritation:

When the appropriate muscles are selected for recording, EMG is sensitive measure for nerve stretch/irritation. EMG is not sensitive to ischemia. EMG activity may be absent during nerve transection.

2. Gauge functional status of a nerve:

Direct electrical stimulation of a nerve is a gold standard for establishing function. For tumor resections, it is common to stimulate the nerve before and after resection to detect changes in nerve conduction.

3. Rule out the presence of nerves embeded within tumor or muscle:

Often times non-neural tissue, such as muscle or tumor, can obscure nerves. Stimulation of tissue can help to differentiate neural versus non-neural. Also, for embedded neural tissue, one can tell the approximate direction and distance to the nerve.

4. Test for medial breach of pedicles following screw placement:

Pedicle screws are frequently used to fuse the spinal column. Direct electrical stimulation of the pedicle screw help to detect medial pedicle cortex breach, which requires repositioning of the screw. Pedicle screw stimulation tests are most common in the lumbar spine. These tests are least sensitive to lateral and superior/inferior breach.

Benefits of Free-Running EMG During Surgery:

  • Relatively noninvasive.
  • Provides immediate feedback when nerves are depolarized by stretch/compression, etc.
  • Informative to anesthesia regarding “reactivity” to painful stimuli.
  • Not markedly affected by inhalational anesthetics.

Limitations of Free-Running EMG During Surgery:

  • Absence of EMG activity is not indicative of functional integrity of nerves.
  • Neuromuscular blocking drugs are contraindicated.
  • Na+ channel blockers are contraindicated.
  • Insensitive to ischemia.

Benefits of Stimulus-Triggered EMG During Surgery:

  • Stimulated EMG is a gold standard for evaluating nerve function.
  • Individual nerves can be identified and differentiated.
  • Can differentiate neural from non-neural tissue.
  • Allows for accurate detection of medial pedicle cortex breach during spinal screw placement.
  • Can provide prognostic information about postop nerve function.

Limitations of Stimulus-Triggered EMG During Surgery:

  • Invasive, can require exposure of a nerve to deliver direct stimulation.
  • Neuromuscular blocking drugs are contraindicated.
  • Na+ channel blockers are contraindicated.
  • Stimulation and recording of Compound Muscle Action Potentials (CMAPs) during assessment of peripheral nerve injury can be misleading.
  • Challenging but not impossible to identify sensory nerves, roots, rootlets.


  1. Delgado, T.E., Buchheit, W.A., Rosenholtz, H.R., and Chrissian, S. (1979).  Intraoperative monitoring of facial muscle evoked responses obtained by intracranial stimulation of the facial nerve: A more accurate technique for facial nerve dissection.  Neurosurgery, 4(5), 418-421.
  2. Hormes, J.T., and Chappuis, J.L. (1993).  Monitoring of lumbosacral nerve roots during spinal instrumentation.  Spine, 18(14), 2059-2062.
  3. Kline, D.G., Hackett, E.R., and May, P.R. (1969).  Evaluation of nerve injuries by evoked potentials and electromyography.  Journal of Neurosurgery, 31, 128-136.
  4. Nelson, K.R., and Vasconez, H.C. (1995).  Nerve transection without neurotonic discharges during intraoperative electromyographic monitoring.  Muscle and Nerve, 18, 236-238.
  5. Nichols, G.S., and Manafov, E. (2012).  Utility of electromyography for nerve root monitoring during spinal surgery.  Journal of Clinical Neurophysiology, 29, 140-148.
  6. Parsons, R.C. (1966).  Electrical stimulation of the facial nerve.  The Laryngoscope, 76(3), 391-406.
  7. Prass, R.I., and Luders, H. (1985).  Constant-current versus constant-voltage stimulation.  Journal of Neurosurgery, 62, 622-623.
  8. Prass, R.I., and Luders, H. (1986).  Acoustic (loudspeaker) facial electromyographic monitoring:  Part 1.  Evoked electromyographic activity during acoustic neuroma resection.  Neurosurgery, 19(3), 392-400.
  9. Prell, J., Rampp, S., Romstöck, J., Fahlbusch, R., and Strauss, C. (2007).  Train time as a quantitative electromyographic parameter for facial nerve function in patients undergoing surgery for vestibular schwannoma.  Journal of Neurosurgery, 106(5), 826-832.
  10. Romstöck J, Strauss C, and Fahlbusch R. (2000). Continuous electromyography monitoring of motor cranial nerves during cerebellopontine angle surgery. Journal of Neurosurgery, 93(4), 586-593.
  11. Schlake, H-P., Goldbrunner, R.H., Milewski, C., Krauss, J., Trautner, H., Behr, R., et al. (2001).  Intra-operative electromyographic monitoring of the lower cranial motor nerves (LCN IX-XII) in skull base surgery.  Clinical Neurology and Neurosurgery, 103(2), 72-82.
  12. Schlake, H-P., Goldbrunner, R.H., Siebert, M., Behr, R., and Roosen, K. (2001).  Intra-operative electromyographic monitoring of extra-ocular motor nerves (Nn III, VI) in skull base surgery.  Acta Neurochirurgica (Wien), 143(3), 251-261.
  13. Strommen, J.A., and Crum, B.A. Intraoperative monitoring with free-running EMG.  In Nuwer, M.R. (ed): Intraoperative Monitoring of Neural Function. Amsterdam: Elsevier, 2008, pp 396-403.
  14. Yingling, C.D., and Gardi, J.N. (1992).  Intraoperative monitoring of facial and cochlear nerves during acoustic neuroma surgery.  Otolaryngologic Clinics of North America, 25(2), 413-448.

2 thoughts on “EMG

  • P April 1, 2016 at 00:50

    There seems to be a growing required use of bipoar EMG montages during spinal IOM over the past few years, the argument being it provides greater specificity and reduces noise and the that there are now more technologies availbale to allow for “more monitoring”. I can see the point of this argument for very specific T-EMG testing like resection of a mass containing or near a specific nerve or localizing a nerve. There also seems to be an improvement in MEP signal acquisition with bipolar EMGs, but this still does not demand that all EMGs should be bipolar. I would like to argue that monopolar, or referential EMGs, are more suited for the demands of IOM when applicable.

    Firstly, specificity is not a huge concern during most spinal IOM, as every individuals anatomy is different, thus are their nerve root-muscle innervation. Nor does it matter much the exact muscles one is recording EMG activity from, just that the EMG activity is recorded, and able to accurately be deemed significant by a properly trained individual, and the surgeon is notified. If that’s true would it not also be true that the increased sensitivity of monopolar EMGs is more beneficial.

    In regards to noise, the adverse impact of the occasional increase in noise associated with monopolar EMG is minimal at best compared to bipolar EMG noise. This in conjunction with the decreased sensitivity and thus smaller recorded amplitude of bipolar EMG activity mitigates its advantage over monopolar EMG signals, as monopolar EMGs amplitudes will to stand out above the possible associated increased noise, thus reducing the advantage of the decreased noise associated with bipolar EMGs.

    On top of this, the use of monopolar EMGs requires less needle electrodes which greatly impacts the effective application of neurodiagnostics in the OR by reducing:
    – harm to pt weather by infection or tape on the patients skin
    – needle stick accidents,
    – surgical delays (setup up time and/or pauses in cases for troubleshooting )
    – incidents of false positives and unneeded stress when its just a technical issue.
    – clutter around the OR table (ie. wires, cable,pods
    – Most important and unavoidable; human error causing more delay, false positives, and a sense of unreliability in IOM.

    Simply put, the OR is not a lab or clinic where you have all day to make sure everything perfect. Reliability and actionable diagnostic information is whats needed most in the OR.

    Thus, although bipolar EMG setup may be optimal in environments without the time or troubleshooting constraints imposed by the OR environment, I feel mono-polar EMG setups can and should be used in most spinal IOM cases to reduce human error, save time, and more realistically align the application of IOM in the OR with the demands of the OR environment. This is especially true and/or when it does not overtly hamper the reliability of MEPs or other potentially more important IOM modalities that significantly benefit from bipolar EMGs.

    Let me know your thoughts or why you believe or know bipolar montages are the preferred method. Your input would be greatly appreciated.

    • Richard Vogel April 10, 2016 at 17:56

      Hi. Thanks for your comment.

      I would certainly agree with you that not all EMG montages need to be bipolar. I’m not a big fan of referential EMG, but I use it when I record from extraocular muscles and it works well for me in that particular scenario. The reason that I use referential recordings is because I’d prefer to place 3 needles in the eye muscles, as opposed to 6. In this scenario, it’s a safety issue.

      In spine surgery, it can be challenging to parse-out EMG from MEPs in many cases. Bipolar recordings for MEPs will allow you to monitor nerve root and peripheral nerve motor function, whereas you lose that “specificity” with referential recordings, and risk false negative results.

      When spine surgery calls for monitoring EMG alone, which I assume would be in most basic lumbosacral fusions, you can probably get away with referential recordings if electrodes are positioned appropriately. I don’t really monitor these procedures, but I’d stick with bipolar recordings if I did. Personally, I don’t see the benefit of referential recordings as you’ve described them.

      From my perspective, harm to the patient is a non-issue. The risk of infection is orders of magnitude higher where they make incision, or place an a-line. The patient receives antiobiotics for this purpose. Infection secondary to placement of IONM electrodes, to my knowledge, is unreported in the literature.

      Needle stick accidents are a concern if you place even one needle. The primary places where needle sticks occur are the head, the hands and the abdomen. So, placing a couple more electrodes in the legs is unlikely to increase the risk in any dramatic fashion. If needles are placed at the appropriate angle, if they are properly secured with tape and stress loops, and if everyone is actively communicating about the location of needles, then that should never happen.

      Surgical delays are also a non-issue. For spine surgery, it should never take longer than 4-5 minutes to set up a patient with electrodes. Any longer than that, and something is wrong, like a missing sense of urgency. Troubleshooting is just a matter of experience. Even the most complex technical issues should be solved/resolved in less than a minute in 2016. Clutter around the OR table is a non-issue. Keep wire neat and don’t let them get tangled in IV lines or other equipment.

      False positives and other forms of human error have little to do with how many electrodes are placed. In fact, placing less electrodes raises the risk of false negative results, which is worse. As for false positives (and false negatives), they are usually secondary to incorrect interpretation, or poor communication about what the data mean.

      I guess, at the end of the day, you are arguing that referential recordings take significantly less time to set up, are safer for the patient, and are less prone to incorrect interpretation. I just disagree, but it’s a matter of opinion and preference. I’m just as fast, just as safe and just as accurate with bipolar recordings.

Comments are closed.

%d bloggers like this: