EMG Techniques - Part I

EMG training Top

Question

I am a physical therapist very interested in developing my skills in electrophysiology and nerve conduction testing. Please point me to the best training program.

Answer

There is an American Society of Electrodiagnostic Technologists (ASET), which is very active in this. They have a website and training opportunities which you can check out at: http://www.aset.org/electroneurodiagnostics.html

Evidence of Re-Innervation in EMG findings Top

Question

In terms of reinnervation findings of EMG, including increased recruitment, instability of MUP, increased Amplitude of MUP (?), etc, which of these features are reliable to predict the true re-innervation in morphology? The reasoning process of EMG physiology is not specific to predict morphological changes.

Answer

One of the first things to appear after denervation is fibrillations and positive waves of course. Following that (in about 2 months) the first signs of reinnervation are an increase in the number of phases of a motor unit, polyphasic potentials (check out: http://www.teleEMG.com/Chapters/jbr120.htm Neurogenic Motor Unit), and increased instability of the MUAP (high jiggle) with decreased recruitment followed by an increase in the amplitude.

Answer 2

Further comment on evidence of reinnervation in addition to concentric EMG findings. The Fiber Density (FD) increases after reinnervation due to collateral sprouting. The increase of FD indicates that muscle fibers increased for the same motor unit. The FD therefore, is a most sensitive method to quantify reinnervation, and therefore, the local organization of motor unit (morphology!). Also, FD corresponds to type grouping in pathology.

EMG Technique (fasciculations) Top

Question

Is there a way to distinguish between benign and malignant fasciculations ELECTROMYOGRAPHICALLY and most of all CLINICALLY? Is it true that rhythmic fasciculations tend to be benign while a singular big thumps tends to be malignant? is it true that brief fasciculations tend to be benign. Is it true that coarse fasciculation are usually benign while the malignant ones are very fine ones? What of all that is true and how it reflects in EMG.

Answer

This is a very common concern and generally no single answer is satisfactory because almost every single rule you make to define a benign versus malignant fasciculation has an exception. Generally speaking though I can tell you that Neurologists rely on far more than the presence (or absence) of fasciculation potentials to make the diagnosis of ALS. If fasciculations become worrisome to a patient, I generally recommend they seek a qualified neurologist's opinion to ease their fears, because the worst thing they can do is to self diagnose them with the disease.

Answer 2

The best measure of benign versus malignant fasciculation potentials comes from the EMG needle exam by the company they keep...in other words... if fibrillation potentials and positive sharp waves are present with large motor unit potentials and polyphasic waveforms with poor recruitment...along with fasciculations...this would tend to suggest a more ominous potential...A good history and physical exam is a priority...remember an EMG exam is only an extension of our physical examination.

Single Fiber Amplifier Settings Top

Question

I am having a minor debate on Amplifier settings for SFEMG. (I believe the Filters were changed on my system.) What are the recommended settings and have there been any recent changes in them? Specifically, I am concerned about the Low Frequency Filter Setting. What would be the upper limit? What is the typical setting?

Answer

The recommended or typical filter settings are LFF 500 Hz and HFF 10 KHz. I may use LFF 1KHz (narrower settings). This LFF setting would reduce the disturbing distant muscle fibers. This setting, however, affects the shape of the single fiber potential. But it should not affect the jitter value or fiber density reading. I am not aware of recent changes. However, if detailed measurement is needed for the shape of action potential, then use more open filter settings LFF 2 Hz and HFF 20 KHz.

Theoretical guidelines vs. practical for thoroughness of EMG studies Top

Question

Having read the guidelines in the online EMG manual for the thoroughness of the study of a muscle (i.e. testing 20-25 units for both spontaneous and voluntary activity, calculating average amplitude phase etc.) I am curious why it seems these guidelines aren't followed more often. Having had a number of EMGs in the past year in both local and university settings and in all cases only 3-4 units (1 insertion 3-4 directions) was sampled. One of these EMGs was overseen by a very renowned name in Electromyography (no names, but odds are you have one of his books on your shelf,) and even then 1 needle insertion per muscle. In a dialog with a doctor who does follow these guidelines, I was told many of the patients he performs EMG on tell him that his study is far more thorough than that done by their own neurologist. I fully understand the reasoning behind the 20-25 MUP guideline if for no other reason than to increase diagnostic accuracy. My questions then are twofold. Do most Electromyographers routinely follow these guidelines or are they followed only when the situation/case dictates. Secondly, how much diagnostic accuracy is there when only 3-4 units per muscle are tested, for example during a MND workup, and no abnormalities are found?

Answer

I agree not all EMGers follow these guidelines, although I do catch myself sampling less frequently in follow-up studies when I am only looking for changes.

Comment

My EMGs have been done locally. I certainly understand your comment about less stringent testing when verifying a change. My main concern is with initial testing where the results are negative, no sign of any disease process. With what degree of certainty can the electromyographer state that the muscle tested is normal when only testing 3-4 units in that muscle? Are they any statistics or figures on how the number of units sampled in a muscle corresponds to diagnostic accuracy rates? I can calculate numbers based on basic probability theory but was curious if any formal studies or accepted statistics existed. Thanks again and kind regards.

Answer

Tough question to answer. I don't know of any studies of predictive statistics (for the number of muscles tested vs. diagnostic accuracy rate) That would be an interesting study. As for number of units tested, naturally if you find the abnormality in the muscle early on (with 3-4 units), you don't need to go any further. But if you don't, the general rule of thumb is to be more thorough to increase your chances.

Answer 2

I agree it would be an interesting study. From a strictly mathematical perspective it should be fairly easy to compute the "odds" of finding abnormalities. This is a lot like the classic probability problem of given that a person throwing darts can hit the bulls eye x% of the time, how many times must they throw the dart to have a given percent chance of getting a bulls eye. In the case of EMG the "x" can be the percentage of units in the muscle involved in the disease process. The number of throws would be the number of units tested and percent chance of getting a bulls eye would correspond to the odds of hitting an involved unit. The equation for this would be: ln(a) = ( ln(i)) / (n)) In this equation the variables are: a - the percent chance of hitting an involved unit expressed as a decimal (i.e. 56% = .56) i - the percentage of units in the muscle involved in the disease process as a decimal n - the number of units sampled in the muscle I have made several assumptions in this equation. 1. The involved units are randomly distributed throughout the muscle. 2. When the needle hits an involved fiber, the pathological signs such as fibs, positive waves, polyphasic units etc. will be visible to the EMGer. 3. The insertion points will be randomly distributed across the muscle. This is not entirely accurate since there would probably only be 4-5 actual insertions through the skin and then testing in multiple directions to get the 20 units. However assuming the insertions were not all done in the same spot, the equation holds. The results of this calculation clearly show why sampling more units can yield more accurate results. Examples: Sampling 20 units in a muscle with 10% involvement gives you a 90% chance of finding an involved unit Sampling 4 units in a muscle with 10% involvement gives you a 56% chance of finding an involved unit. Sampling 20 units in a muscle with 25% involvement gives you a 93% chance of finding an involved unit. Sampling 4 units in a muscle with 25% involvement gives you a 70% chance of finding an involved unit. Sampling 20 units in a muscle with 50% involvement gives you a 97% chance of finding an involved unit. Sampling 4 units in a muscle with 50% involvement gives you an 84% chance of finding an involved unit. So sampling more units in a muscle greatly increases the chances of finding an involved unit, especially early in the disease process. Kind regards

Answer 3

I read the interesting discussion on this point. I believe that few additional factors may influence the number of motor unit tested. Firstly, the experience of the neurologist or electromyographer. Secondly, the clinical examination of the patient. And finally, the degree of patient's cooperation during sampling. It is also, basically, the same argument could be applied to the number of F-waves should be obtained for shortest latency. My kind regards,

Answer 4

Very interesting subject. About 4 years ago I was in Italy / Milano/. At the very famous S.R. Hospital I saw the following practice. The electromyographer there does not catch even one potential at single muscle. They say to the patient to do a slight effort, then observe moving potentials (do not record), after e period of few minutes they made conclusions e.g. MUP with growth of amplitude and area. I asked them how long of a period is needed to become proficient in this method: - 3 months