The goal is to become familiar with some of the techniques of recording gross potentials in the auditory system. This involves the use of (relatively) large electrodes, usually made out of silver wire because of its low resistance and resistance to corrosion and more importantly, it can be chlorided so that it can either accept or lend an electron in passing current.

• Why might this be important?

Surgery: A chinchilla will be prepared for recording: this involves anesthetization with sodium pentobarbital, dose rate of 75mg/kg of body weight. Additional doses are given at roughly 10% of the anesthetic dose and determined by whether the animal responds to a paw pinch. Its respiratory pattern is also watched (if the rate increases then give an additional dose. If it is irregular/ intermittent, no additional dose as the animal could be expiring.). A tracheotomy is performed as this reduces the likelihood of the respiratory airway becoming obstructed. The animal's temperature is kept at 37°C through the use of a rectal probe and a homeothermic blanket on which the animal rests.

The animal's skull is cleaned in order to accept a head mount to which a post is attached. The bulla is then exposed and an opening is made so that the round window can be visualized. A wire is then positioned so that it contacts the round window. The wire is cemented to the bulla. A hypodermic needle serves as a reference electrode and is positioned in a jaw muscle. This goes to a preamplifier which acts as a buffer and also reduces the noise that would otherwise be present at the input to the computer A/D system. An opening is also made in the ventral bulla so that the tensor tympani can be cut. The opening is then sealed after a vent tube is placed in it.

• What is the function of the vent?

Have someone speak into the open meatus and observe the signal on an oscilloscope.

• Why is the signal called a microphonic?

An earphone is coupled to the meatus and a probe tube microphone is inserted. The phone is calibrated using a program called Neucal. This program presents short tones at increasing frequencies, samples the output of a microphone sensing the pressure at the eardrum of the subject, averages the response, and after all the frequencies have been presented, it calculates the acoustic transfer function (The microphone and probe tube have been previously calibrated so that dB per Volts at the input to the A/D is known.) The calibration is to be made from 100 to 20,000 Hz in 50 Hz steps. The resulting amplitude plot corresponds to the sound pressure level (dB re 20 mPa) for 0 dB attenuation. The attenuation covers a 0 to 127 dB range in 1 dB steps. The signal generator covers 0 to 65 kHz.

A plot of the calibration is to be made. This calibration will also be stored in the computer for use in the data collection process (so the amplitude can be corrected for the phone characteristic).

CM Potentials.

One of the first potentials ever recorded in the auditory system was the cochlear microphonic (CM). This is the integrated output of some portion of the cochlea generated primarily by the OHCs. It has no threshold. It will appear in the signal of other recordings just because it can be rather large - millivolts. At times you may have to take measures to eliminate it from recordings when you are not interested in it.

One can use the EP program to record CM and EPs over a response space that you specify: Try 500 to 20000Hz in 1000Hz increments: 0 to 90 dB SPL in 10 dB increments. You will obtain a plot after every frequency that you should save and study. From the menu that is displayed after each freq., type 5 to print the graph on the laser printer, and then select 8 to also save the data for later retrieval and spectral analysis. Record the dataset number which is automatically assigned by the program - you will need it later to retrieve your data.

Since the CM follows the motion of the basilar membrane (remember Dallos' hair cell studies), the waveform will invert if you invert the polarity of the acoustic signal. This fact has been used when you want to focus on N1 potentials. That is, if you alternate the polarity of the acoustic signal the CM will cancel upon averaging while the N1 potential will become more prominent. Try it. Use 2000 Hz and 70 dB SPL. (Hint: Answer YES to the 'alternate phase' option).

You can examine the spectral characteristics of the previously
saved waveforms as follows (while logged in on MVF):

$ STATPK
> EM EPSPEC

the macro EPSPEC will ask you for the dataset number (recorded earlier during data collection). It will then plot (on the laser printer) a graph that shows the waveforms in both time and frequency domains. Do these plots for selected datasets only.

N1 Potentials.

One commonly used method of monitoring the viability of the cochlea during an experiment is to determine the N1 threshold using visual detection criteria. The N1 is the integrative signal produced by the auditory nerve due to a synchronous firing in response to an acoustic signal. A click works best but has a limitation since it excites the entire cochlea, hence all auditory nerve fibers. The modification that is used in order to ascertain the function of separate regions of the cochlea is to shape a short tone pulse (at different frequencies) and present it enough times to obtain a reliable average. At sufficiently low levels only a relatively restricted region of the cochlea is excited, hence, one can obtain the threshold in a frequency specific manner. The threshold data becomes a reference for future use: if after further manipulation or just the passage of time, the thresholds have risen then one makes a note of it and if the change is larger than can be accepted one has to terminate the experiment.

Try to obtain the N1 response at 500 Hz and 1000 Hz. Vary the intensity from 30 to 80 dB SPL. Note that it is more difficult to obtain N1 at low frequencies. Why?

How is the response different from the response to higher frequencies?

How does the N 1 latency vary as a function of frequency and SPL?

The last person to try the experiment: make an opening in the tympanic membrane about 2mm in diameter. Repeat the measurement of the N1 threshold curve. What changes occurred?

• What pathologies could be studied with these techniques?

• What alternate measurement techniques could/are used to determine the nature of middle and/or inner ear pathologies?

ABR

ABR (auditory brainstem recording) is a far field potential: the recording electrodes are distant from the generators in the brainstem. The peaks in the resulting waveform are numbered using the Roman numerals. In particular, wave V corresponds to a source in or near the IC.

AEPs (auditory evoked potentials) were recorded in the 1930's. The recordings were largest when the electrodes were on the vertex. There are late potentials that occur with a latency of 50 to 200ms.

AEPs result from synchronized neural discharge patterns and graded post-synaptic potentials and are of clinical significance, e.g., detection of acoustic tumors.

CM: primary receptor potential. There is no threshold for CM recordings.

SP: 2^nd receptor potential. Summating potential. Some believe they can be used in the detection of certain end organ disorders and that they originate from OHCs.

Latency: SP & nerve: 0-2ms.

Nerve and brainstem: 2-10ms

Thalamus & cortical: 50-300ms.

Late association cortex: >300 ms.

See the tables and illustrations of ABRs attached. Remember that the filter settings may be crucial for comparision with previously gathered/published data.

Record ABRs (averaged brainstem responses) from the vertex of the skull. Using a teflon-coated silver wire, place one end on the skull along the midline. The reference can be almost anywhere (on the animal). The filters should be set to: 3 - 3000 Hz.

• What problems might arise if the reference is on the torso?

The stimulus is the click (impulse). These potentials are quite small so that amplifications of 50,000 to 100, 000 are routinely used. You will need to insert a second amplifier in your signal line before it is sampled by the A/D. The goal is to utilize the full dynamic range of the A/D system that can cover ± 5V. So the output of the amplifier should cover the same range.

Begin with 70 dB SPL & average 1000 click responses.

If this doesn't yield anything try a higher SPL. Another variable is the number of averages. Try 5000. This of course will take longer. The Signal/noise ratio improves by a factor of 2 for every increase of 4 in the number of averages.

Remember: there could be artifacts averaged in if proper care is not exercised.

• How does this response compare to the ABR for a human?

What might account for some of the differences?

Include an example of the ABR in your report.

Label the neural generators (probable sources) for the different waveforms.

Hand in: A set of curves with N1 responses indicated. Create a TH vs Frequency plot for CM and N1. Indicate the threshold of CM and N1 vs log (frequency).

Write up this experiment as a short report. Do this individually. Use a word processor.

Provide an Introduction.

Methods. (describe the experimental setup. E.g. Filter settings, amplifiers, A/D, connections, software, etc.)

Results.

Discussion.

Addendum.

Repeat the questions of this exercise and give your answers in the report.