Written by Travis M. Moore
Last edited 19-Jul-2020
Congratulations, you are now at a point where you can begin learning specific electrophysiological tests! The next several lessons will cover various auditory evoked potentials, including: their proposed site of origin, recording parameters, stimulus parameters, waveform morphology, and clinical applications. We will organize our survey of clinical assessments by starting with tests that examine more peripheral structures, and ending with tests that record the most central potentials from the cortex itself.
The cochlea is the most peripheral structure in the auditory system that gives rise to electrical potentials. Believe it or not, we can measure the tiny electrical activity from the cochlea in response to sound. The audiological test for doing so is called electrocochleography, abbreviated ECochG. Electrocochleography is also sensitive to the firing of auditory nerve fibers (first order neurons) activated by inner hair cells (IHCs).
Let's be clear though, the ECochG potentials that we measure in
clinic arise from the OHCs and the auditory nerve fibers (ANFs). While
the inner hair cells cause ANFs to depolarize, we're
not actually measuring IHC potentials here.
The output of ECochG is called an electrocochleogram, and is unfortunately also abbreviated ECochG. (You could also see ECochGm.) You'll know from context clues which word is meant.
Electrocochleography produces three waveforms: two from the OHCs and one from the ANFs. The potentials measured from the OHCs are called the cochlear microphonic (CM) and the summating potential (SP). The OHC potentials are from a class of potentials known as receptor potentials, which means the CM and SP arise from sensory cells. In other words, they show sensory transduction happening (i.e., OHCs depolarizing). Because the activity of receptor potentials is tied to sensation, their waveforms tend to "follow" the shape of the stimulus, rather than produce peaks and troughs related to more abstract neural firing like most potentials. The potential measured from the ANFs is called the compound action potential (CAP), and represents the initial activity of the auditory nerve to sound.
|Cochlear Microphonic||CM||Outer hair cells|
|Summating Potential||SP||Outer hair cells|
|Compound Action Potential||CAP||Auditory nerve fibers|
If we present a 250-Hz tone burst, we should see a roughly sinusoidal-looking CM on the computer screen at 250 cycles per second. Not convinced? Let's think about why this happens. A 250-Hz sine wave enters the cochlea, containing a series of peaks and troughs. At each peak, the basilar membrane moves, and OHCs fire. Since that's what we're recording, well see a change in voltage. The the sine wave drops into a trough, the basilar membrane moves again, and OHCs fire again. We see another change (i.e., a drop) in voltage. So really it makes sense that we see a waveform that follows the stimulus if we're recording from the sensory cells involved in transduction. See Figure 1 below for a look at the CM.
In fact, the waveforms recorded from auditory neurons mimic the stimuli so closely, if you play the recorded electrical waveforms through a speaker, you can reproduce them! Listen to the audio clip below from Dr. Nina Kraus' lab, which starts playing recorded music, then switches to neural output recorded from the brain(stem). The evoked potential used for this recording was not the CM, but the frequency-following response (related in principle, but from the brainstem).
If you don't think that's cool, you're too hard to impress. Let's make an observation based on what we just heard. You know that far-field neural recordings are messy due to unwanted electrical noise. Those recordings were just played through a speaker. Notice all the extra acoustic noise in the signal when the sound clip switches to the electrophysiological recording? Now instead of seeing noise in recorded potentials, you're hearing it played as sound. Again, if you don't think that's cool, you need to recalibrate.
The SP is a little less exciting. Instead of following the peaks and troughs (i.e., fine structure) of the stimulus, the SP shows a change in voltage when a stimulus begins that is maintained until the stimulus ends. It's sort of like an "on/off" indicator. If there is no stimulus, the SP is at baseline, if a stimulus is played, we'll know when and for how long due to the change in voltage from the SP. Figure 3 below shows the brief SP in response to a click. Note that the CAP immediately follows the cochlear activity indicated by the SP.
Are you starting to sense a problem with recording the CM and SP? If the CM mimics the stimulus and the SP is "on" during the stimulus, that means both potentials occur at the same time. And yes, that can be problematic for obvious reasons. The waveforms we record are messy enough without having to deal with simultaneous signals. What typically happens is the CM "rides" on the SP (Figure 4). It turns out the CM isn't very diagnostically useful at the time of this writing, so ideally we could just get rid of it. Luckily, there is a simple but elegant solution. Can you guess what it is?
Remember the CM mimics the acoustic stimulus. And how do we cancel out acoustic stimuli? Just like the differential amplifier, and as discussed in the lesson on additive synthesis, if we combine two waves that are 180 degrees out of phase ("flipped"), they cancel each other out. If we apply this logic to our ECochG stimulus, we can alternate presenting clicks that start with a peak (i.e., rarefaction) and clicks that start with a trough (i.e. condensation). Averaging over many sweeps, the alternating polarities cancel out the CM. That is, OHCs' response to the peaks and troughs, which are positive and negative, average to zero. But, the SP responds the same way (on or off) regardless of the polarity, so it is preserved. Take a minute to bask in the intersection of so many principles you have learned from various fields (e.g., acoustics, neurophysiology, electromagnetism) and realize how complex but simple this concept is.
The CAP is more straightforward than the cochlear potentials. We present a brief click, thousands of ANFs depolarize in synchrony, and we record the summed voltage that reaches the scalp. If you're wondering why this is referred to as the compound action potential (and not just action potential), it's to highlight the fact that thousands of fibers contribute to the response we see on the computer screen. This is the type of evoked potential we have been discussing all along, so there's not much else to say.
Now that you know what the waveforms of the ECochG look like, it's time to learn how to quantify them. Recall that the two main pieces of information from evoked potential testing are (1) latency and (2) amplitude. The expected latency of the CAP is around 1 to 2 ms (the same) as Wave I of the ABR). So in the sea of squiggles you're likely to encounter, start your search at 1 ms. The SP is cochlear in origin (almost entirely from the OHCs), so it makes sense the SP should appear immediately before the CAP. In fact, it typically appears on the left-hand slope of the CAP (e.g. Figure 3). What you're seeing is the cochlea respond, then the auditory nerve.
There are two ways to report ECochG amplitude: peak-to-trough and using a prestimulus baseline. The latter seems to be used most often so we'll focus there, but both methods are shown in Figure 6. Remember that a prestimulus baseline involves recording a little "quiet" time before the stimulus is presented. The idea is that there won't be any huge voltage fluctuations if the patient is sitting quietly. Simply choose a point that is clearly before the SP (in a flat area if possible), and use that as your reference point for both the SP and the CAP amplitude. That means in order to find the SP amplitude, you would measure the change in voltage from the baseline, to the tough of the SP. To find the CAP amplitude, you would measure the change in voltage from the baseline to the trough of the CAP.
Recall that the inner ear is housed in the petrous part of the
temporal bone, which is the densest bone in the body. (It's so
dense it's often what CSI techs use to extract DNA when trying to
identify a body.
|Site||Electrode Type||Typical Amplitude6|
|External auditory meatus||Foil electrode / tiptrode||1.5 μV|
|Tympanic membrane||TM electrode||3 μV|
|Promontory||Transtympanic / needle electrode||15 μV|
When we record from the EAM, we're using the same principles as when we record from the scalp, we're just using the skin inside the ear canal. Now, there's no need to reinvent the wheel when deciding how to get an electrode in the EAM; we already have something that fits snugly in there: the foam tip of an insert earphone. We'll need an foam insert anyway to deliver the stimulus. Because foam doesn't conduct electricity, these foam tips are wrapped in thin gold foil (recall that gold is an excellent conductor. The idea is that the insert will expand and press the metal foil against the EAM (as well as pass sound). The electrode leads clip onto the gold-wrapped stem of the insert with tiny alligator clips. You can see what these gold foil electrodes (sometimes called tiptrodes) look like in Figure 7.
Gold foil electrodes seem to be the most common in U.S. clinics because they are non-invasive. You'll need to slather the foil with an electrolyte gel (similar to the stuff in the foam on a regular electrode), so it will feel damp, but otherwise no more inconvenient than a typical foam tip. The tradeoff is that recordings look fairly noisy and low-amplitude.
TM electrodes are just what they sound like: modified electrodes that rest against the TM. If you're wondering whether that requires local anaesthetic, the answer is no. It's not the most comfortable thing in the world, and placing them takes practice, but they're livable. The increased invasiveness gets us closer to the cochlea, and therefore produces a larger, cleaner recording. Take a look at Figure 8 to see what a TM electrode looks like.
There's not really much to see here - it consists of a length of wire that is fed down the ear canal, and ends in a small blob of foam and electrode gel. The opposite end is business as usual and provides a connector to plug the electrode into the terminal box.
The answer to your other question is "yes," audiologists can and do place these electrodes. You need to use an otoscope so you can guide the TM electrode down the ear canal and make sure the end is resting nicely on the TM. These electrodes are quite expensive for a disposable tool, so these are relatively rare in clinics. Also that whole placing something on the eardrum thing.
The most invasive ECochG site in humans is the cochlear promontory. This site can be reached using a transtympanic electrode (trans- = through). This electrode is also called a needle electrode, which describes it very well (see Figure 9). Audiologists do NOT place this type of electrode, and it does require a local anaesthetic. Figure 10 illustrates how the needle is placed against the promontory.
The tables below provide the electrode montage, amplifier settings, and stimulus parameters typically used for ECochG with gold foil and Ag/AgCl electrodes.
|Non-inverting (active)||Tiptrode in test ear|
|Inverting (reference)||Contralateral mastoid|
|Ground (common)||Fpz (low forehead)|
|High-pass filter||10 Hz|
|Low-pass filter||1500 Hz|
|Intensity||90 dB nHL|
This section follows the Frontiers in Neuroscience review paper
by Gibbons (2017).
Before less invasive tests became available, ECochG provided an objective method of assessing hearing function in children and other populations unable to respond behaviorally. Keep in mind the only information gained from ECochG for assessing hearing is the presence/absence of the CAP. So ECochG is not the same as a behavioral hearing test, but rather a test of OHC and ANF function. Futhermore, because ECochG is so difficult to measure using noninvasive electrodes (see Electrodes subsection below), the only way to find the CAP threshold is to record from an electrode touching the bony labyrinth.
Generally speaking, auditory neuropathy spectrum disorder (ANSD) is
characterized by intact OHCs, but damaged ANFs and/or auditory
Endolymphatic hydrops (ELH) refers to increased pressure in the membranous
labyrinth of the inner ear due to an excess of fluid. That extra
fluid pressure causes changes in how the basilar membrane and vestibular
organs operate, which can lead to a variety of symptoms. Most commonly,
ELH is associated with Ménière's disease (MD), but the two are not
synonymous. ECochG testing typically reveals a larger than normal
response (e.g., SP/AP ratio) in the presence of hydrops, and is
therefore sometimes ordered when attempting to diagnose MD (which is
not easy). What you need to keep in mind is that (1) ELH does not equal
MD, and (2) the sensitivity of click ECochG to hydrops is not that
Figure 11 shows the results of an example patient. The left panel reveals a larger than normal SP, leading to a larger SP/AP ratio (i.e., greater than 0.3). The right panel shows a normal SP amplitude. Thus, these results suggest the patient has ELH in the left ear, but not the right.
Let's first define intraoperative monitoring. The prefix intra- in Latin means "within," so intraoperative means something that happens during surgery. In this case, we are monitoring responses from the cochlea (typically the CM) during a surgical procedure. Before this technology was available, surgeons could only take their best guess in trying to avoid damaging the inner ear/auditory nerve. With the advent of ECochG, audiologists can now continuously record the CM while the surgeon performs surgery. If the amplitude of the CM begins to decrease, the audiologist can alert the surgeon they are getting too close to important auditory structures and to try another approach. If the CM completely disappears, then damage has occurred and the audiologist works with the surgeon to try and restore function. These types of monitoring procedures have greatly improved hearing preservation during surgery, and are legally required in some areas.
Common procedures that can be monitored with ECochG:
The American Speech-Language-Hearing Association (ASHA) has a good page about short latency auditory evoked potentials, including general principles, ECochG, and ABR.