Yost Chapter 6: The outer and middle ears pp. 65 -79 [2 February 2004]

 

Figure 1 gives the domain of the next section, Chapters 6, 7, and 8. By the end of this section you should be able to go to the blackboard and draw out a reasonable sketch of the peripheral auditory system, and describe the function of each of its parts. Note that (of course) this is a human ear. Other mammals share these features in various proportions, mostly controlled by body weight: but there are major evolutionary changes from reptiles and birds to humans, some of which we will discuss in light of their special adaptations.

STRUCTURE OF THE OUTER EAR.

Yost describes briefly the pinna as the “visible part of the ear,” which in humans seems like an esthetic appendage useful only in concert with the nose for holding spectacles: quite different from the ears of, say elephants or rabbits, which both “catch the sound” and also are used to cool the body. In contrast the human ear has no useful muscles and cannot be moved without the intercession of the hand (people who can wiggle their ears are thought to wiggle their scalp). It has distinctive ridges and whorls, each of which has a special and difficult to remember name which are important largely to plastic surgeons. The visible part of the ear is composed of cartilege and skin, save for the ear lobe. The rounded outer surface is the "helix" and the rounded curved portion separated from the helix by a small valley is called the "anti-helix". The “concha” or cave describes the inside cavity formed by the pinna, and its function is to take the sound waves captured by the pinna and funnel them into the ear canal. The little projection at the front of the concha is called the "tragus" and is though to serve a protective function. Why tragus? I looked in the OED to find out that tragus means "goat" in Greek, and you might have noticed that some grandfatherly types have hairs growing on their tragus: hence tragus is like a little goatee, that some men grow on their chin, where it is definitely more attractive. The hairs may keep out dust though, and perhaps small insects. The funnel is the “external auditory meatus” and the ear canal is the external auditory canal (because there is also an internal auditory meatus and canal, through which nerves and blood vessels enter and leave the brain to and from the ear). Yost writes that the canal is 2 to 3 cm in length, and about 5 to 7 mm in diameter, with a cartilaginous and hairy outer third, and a bony inner two-thirds, which ends in the tympanic membrane. The wax and the hairs serve a protective function against dust, dirt, and small insects. One easily remedied form of hearing loss is an excessive build-up of wax blocking the ear canal (this is one form of “conduction deafness”). The outer ear obviously funnels sound into the ear canal, but an unsuspected function is its modifying the sound spectrum depending on the direction of a complex sound, and thus allowing us to locate sounds in the vertical plane. The depth of the ear canal, as we know, is important for its particular resonant frequency which somewhat resembles the speech spectrum.

STRUCTURE OF THE MIDDLE EAR

Yost begins with the tympanic membrane, which he describes as “one boundary of the large cavity known as the middle ear cavity, or tympanum” most of which is enclosed between the tympanic membrane and the outer wall of the inner ear, with a little superior portion called the epitympanum in which the hinge separating the malleus and the incus sits, and a long tube, the eustachian tube, which connects to the “nasopharynx”, or the nasal cavity. The volume of the tympanum is about 2 cubic centimeters in humans, and the eustachian tube is 35 to 38 mm long. This tube is of course important in equalizing air pressure on both sides of the tympanic membrane: in infants it is immature and I think it is oriented in a way that tends to attract food stuffs and related bacteria, hence infants and some older children have lots of middle ear infections. Note that the tympanic membrane is cone shaped, about 55 to 90 square mm in area, again in humans; the only real significance of this number is in its relation to the area of the oval window, which we find out on page 68 is 3.2 square mm in area, so one would think the mechanical advantage has a range of maybe 20:1 or 30:1 Yost takes the smaller one and calculates an increase of 17 to 1. Why does he take the smaller one? Because later on he writes that at high input levels only part of the tympanic membrane vibrates (page 68).

Pickles does the same calculation but for the cat and comes out with a figure of 35 to 1 -- it seems that the cat has a tympanic membrane almost the size of the human at 42 square mm, and a oval window of just 1.2 square mm. Does the cat hear better than we do? And if so, is it because of this difference in the hydraulics of the middle ear? The answer to the first question is “Yes it does” -- several studies show that it can hear about -10 to -20 dB in the range of 1 to 8 or 10 kHz, compared to the humans 0 dB to maybe -4 dB. One could guess that maybe 6 dB comes from its ability to double the pressure at the oval window. The tympanic membrane is made up of an interwoven mat of elastic fibers, and at the head of its cone (the “umbo” -- in Latin, the point on a shield) it is attached to the mallet, or hammer, or “malleus”, the first of the three bones or ossicles of the middle ear, with the attachment site being the “manubrium” (or handle -- note “mano” for hand in Latin and Spanish). Then the manubrium in connected to the “incus”, which descends out of the epitympanum and connects to the “stapes” or stirrup, which connects to the oval window. The stapes is the smallest bone in the body, yet is divided into 4 parts: the head (“capitum”), two arms or struts (the “crura”) and a footplate. Around the footplate runs the annular ligament which holds the footplate to the oval window (and is often a place where a bony growth begins and leads to a sort of arthritis and another form of conduction deafness.) The bones are held in place by other ligaments (see Figure 6.4) and two muscles, one, the tensor tympani, which is attached to the handle of the malleus, and the other, the stapedial, attached to the head of the stapes. The tensor tympani is about 25 mm long, and the stapedius is 6 mm long -- it is the smallest muscle that pulls on the smallest bone in the body. They are controlled through the cranial nerves 5 and 7, the trigeminal and the facial. The photographs given in Figure 6.3 a and b are quite remarkable. Because the middle ear is an enclosed space it also has its own resonance, which again roughly matches the speech frequencies. A major function of the middle ear is to amplify the pressure at the inner ear so as to make up for the impedance mismatch between air and saltwater, as we know. But also, because of these small muscles and their neural control it is possible to stiffen the system and make it less sensitive to low frequencies (again, think of the effect of an increase in stiffness on resonance), and thus reduce masking and also provide protection against acoustic overload. FUNCTION OF THE OUTER EAR

Yost begins with a story on changes in transmission in amplitude and phase spectra from one place to the next, a graph of the input vs. the output being called a “transfer function” (of which we will find many). Yost shows in Figure 6.5 “Head related transfer functions” for amplitude and time differences (called Interaural Level and Interaural Time Differences, ILD and ITD) across frequency. Note that this is at the entrance to the ear canal. There is a small increase for mid frequencies at the near ear, and the sharp decrement beginning at about 4 kHz for the far ear, with a difference in intensity of a few dB beginning at around 1 to 2 kHz. Note also the time differences, which are on the order of 700 to 800 microseconds, a tiny bit longer for low vs. high frequencies. Note also that this is a difference in onset time, but it also is a difference in phase: but for a constant time difference the phase difference is going to vary with frequency. The nervous system clearly uses phase differences, but you can see that this will lead to a computational nightmare for a complex signal. In part this is accomplished in the midbrain binaural centers which are arranged 2-dimensionally in phase and frequency. The pinna also reflect high frequencies differently from low frequencies which provide small echoes to the nervous system, which then uses then to add to the perception of location.

Figure 6.6 gives the transfer function for the different parts of the external ear as measured at the tympanic membrane, and the critical feature is that there is a peak in the region of 2 to 3 kHz for the total of the concha and the external ear canal. The peak is large and approaches that of the audiogram in its strength. Apart from providing a special resonance, it is usually thought that the outer ear keeps dirt and bugs away from the middle ear and maintains the tympanic membrane under relatively homogeneous conditions of humidity and temperature. But also the pinna especially changes the shape of complex wave fronts according to their direction, which is useful in locating objects varying in the vertical plane.

FUNCTION OF THE MIDDLE EAR.

Yost points out that there are three ways for sound to reach the inner ear -- through vibration of the skull (either from air, or from a blow to the head -- or from talking), through the air in the middle ear cavity, and through the ossicular chain. Obviously this third one is preferable. He goes on to talk about the problem of air molecules acting on fluid (not very well) as propagation of the sound wave would be impeded (hence “impedance”): hence the value of the ossicular chain and the pressure exerted by leverage and hydraulics, namely, the pressure = F/A formula from before. And why is a 17-fold increase in pressure equal to 15 dB? Is it because 20 x log(17) = 15? Well actually it doesn’t: 20 x log(17) = 25. And then there is the leverage factor of the ossicles, which is about 1.3, and then some other sort of fudge factor which is an estimate of the effect of buckling the cone, about 2: which gives a pressure change of 17 X 1.3 X 2 = 44 , and 20 x log (44) = 33 dB: hooray! (See middle of page 74). But this is not true of all frequencies, because the bones have mass and the ligaments and muscles provide stiffness. In Figure 6.8 we see the transfer function of the middle ear providing a peak around 1 kHz, and that the maximum gain is about 25 dB. (In the text Yost says the maximum gain he sees in Figure 6.8 is over 30 dB in the region of 2500 Hz. I do not know why he comes to this conclusion.)

Yost then provides some information on the middle ear reflex in which the muscles tense (the stapedius muscle has the greater effect I think) at a threshold value of about 70 to 80 dB in humans, primarily for low frequencies (below 2 kHz): why this? Because the stiffness increases and so the resonance frequency goes up. Yost says reflex action takes 10 ms, but this is not to full protection. The latency of the first sign of reflex action is 10 ms, but full protection may take much longer, up to 150 ms. For this reason it protects against prolonged sounds, but not short impact sounds, solitary gun shots for example, though it would protect against subsequent repetitive gunfire. However, it seems unlikely that gunfire was the evolutionary spur for the middle ear reflex. The function of the reflex is protection, possibly, but also to keep the ossicles together and thus prevent distortion (the addition of new frequencies not in the input). Some authors argue the real function of the reflex is to prevent our speech from masking our hearing. Yost also goes into a very subtle point on page 76 (Figure 6.9) describing a switch in the action of the stapes from “in and out” to “rocking” at high intensities. When it rocks back and forth the total motion transmitted to the inner ear is minimal, which Yost sees as another protective mechanism for very high sounds -- possibly those that occur just after one is hit by lightning. His point is that this protective influence is not great, given the number of people who show severe noise induced hearing impairment. But of course, evolution did not prepare us for the Sony Walkman or the subway train. A recent analysis by David Smith (ARO 2002) shows that sounds intense enough to produce hearing loss do not occur in natural conditions (save for nearby lightening strikes), but there is some controversy about this.

MIDDLE EAR IMPEDANCES

In Chapter 3 note that impedance is made up of resistance and reactance, and reactance has two components, related to stiffness and to mass. Resistance does not vary with frequency while both forms of reactance do. Resistance in the middle ear is mostly at the oval window; mass is related to the weight of the bones (so this component of reactance increases with frequency) and stiffness is related to the springiness of the ligaments and muscles (and this decreases with frequency). Stiffness provides high reactance for low frequencies, which is why tightening the muscles primarily depresses low frequencies; the mass produces high reactance which depresses high frequencies; and resistance is most important in the middle frequencies.

Note that “compliance” is the opposite of stiffness. Yost goes on to describe the “electro-acoustic impedance bridge” which measures the compliance of the middle ear, by measuring the degree to which the tympanic membrane reflex reflects sound back out of the ear canal. It uses a low frequency tone for the reflection, which is affected mostly by stiffness, rather than mass, and so it is able to show the effect of the reflex stiffening. An increase in muscle tension increases stiffness and lowers compliance -- more low frequency is reflected out. This technique measures problems in the middle ear caused by the increase in pressure due to infection, it can determine whether there are holes in the tympanic membrane, and it can determine whether the middle ear reflex is operating properly, thus providing some information on neural activities. It is a standard tool in the audiology clinic.

Air transmission (in an “open” middle ear) and bone transmission (transmission through the skull) are much less effective than the middle ear transmission described above, for obvious reasons, including the fact that air transmission through the middle ear must affect both the round window and the oval window simultaneously. Bone conduction is not very valuable under normal listening conditions, but it does make for a useful diagnostic test -- if a person can hear through bone transmission but has an impaired normal audiogram, then there must be a conduction problem in either the outer or the middle ear, because the inner ear is reasonable sensitive to pressure changes transmitted through the skull. This form of “conduction deafness” in chronic cases is remedied by antibiotics or by inserting tubes through the tympanic membrane (for middle ear infections) or surgery (for rebuilding arthritic intratympanic bones and their connections). Hearing aids really help conduction deafness.