Lilly, David J.1; Margolis, Robert H.2
The 2012 Eriksholm Workshop focused on some of the most recent acoustic measurements within the occluded, human external auditory meatus (EAM). The goal of this introduction is to provide an overview of basic and clinical EAM measurements that evolved in the 20th century and some relations between these measurements and wideband acoustic absorbance.
For nearly all of these measurements, an acoustic signal is introduced into the closed EAM and a microphone in the same space is used to sample the sound pressure generated. The resultant voltage at the output terminals of that microphone then is processed to estimate acoustic conditions within the EAM and at the lateral surface of the tympanic membrane (TM).
Two groups of investigators contributed the bulk of experimental data in the early years. The first group was concerned with telephone communication. The second group included basic scientists and clinicians who were interested in middle ear mechanics and in the differential diagnosis of middle ear disease.
STANDARDIZATION AND CALIBRATION OF ACOUSTIC TRANSDUCERS
To standardize and calibrate the output transducer of a telephone (the “receiver”), the first group of investigators needed to measure the acoustic “load” offered by an average adult ear. With these data in hand, they then could design an “artificial-ear coupler” that could be used during the manufacture and the standardization of telephone receivers. Most readers of this supplement, of course, are familiar with an array of acoustic couplers that are used on a regular basis today to evaluate and calibrate acoustic transducers.
It should come as no surprise to learn that much of the early research in this area was published by workers in telephony. Inglis et al. (1932) published a germinal report on the acoustic load offered by the closed EAM to a telephone receiver. They noted that “the human ear is limited in its utility generally to relative or comparative measurements and is not well adapted for measurement of absolute pressure or velocity.” In consequence, the work outlined in their article was directed toward the development of an “artificial ear” that would present acoustic characteristics that reflected the acoustic load of a normal adult ear more accurately than simple closed cavities. Figure 1 has been reproduced from Inglis et al. The top panel (A) is a cross-sectional view of an adult human pinna occluded by the cap of a telephone receiver. The lower panel (B) depicts their artificial-ear coupler, again, covered with the cap of a telephone receiver. It is interesting to note that Inglis et al. have used the symbols Z, Z1, Z2, and Z3 to indicate four components of the total acoustic load. Today, we would identify these acoustic impedances with an A subscript (ZA). Heaviside (1886) probably was the first to use the term Impedance in the analysis of an alternating-current electric network. Steinmetz (1893) introduced the use of complex numbers as a tool to identify both the resistive and the reactive part of any impedance. The utility of the impedance concept was enhanced when Webster (1919) suggested that the term also could be extended to mechanical and acoustical circuits. He wrote that “whenever we have permanent vibrations of a single given frequency … the notation of impedance is valuable in replacing all of the quantities involved in the reactions of the system by a single complex number.”
MEASUREMENTS OF STATIC ACOUSTIC IMPEDANCE WITHIN THE EAM
The second group of investigators who studied acoustic conditions within the EAM and at the lateral surface of the TM was concerned initially with static acoustic impedance (ZA). Their measurements were made with ambient air pressure in the ear canal and with the middle ear muscles in a state of normal tonus. The history of static, aural acoustic-impedance measurements, relations between these measurements and anatomical structures, and some effects of common middle ear diseases have been reviewed in an earlier publication (Lilly 1973). In that publication we provide also a review of: (1) Instrumentation that has been used to measure aural ZA; (2) standard methods for presenting ZA data; (3) equations for converting these data from polar to rectangular form and vice versa; and (4) a summary of mean ZA data for “normal” middle ears, for patients with clinical otosclerosis and for patients with interruption of the ossicular chain over the frequency range from 125 Hz through 1000 Hz.
In general, for patients with clinical otosclerosis, the real part of the acoustic impedance at the TM (ZTM), the acoustic resistance (RTM), falls within the normal range. The imaginary part of the measured ZTM, the acoustic reactance (XTM), however, is very much higher than normal. These findings are displayed graphically in Figure 2. This figure has been reproduced from Lilly (1973). For patients with ossicular discontinuity, the measured RTM and the acoustic reactance (XTM) both are much lower than normal. These findings also are displayed graphically in Figure 2.
Pattern recognition has been used extensively in clinical audiology. Inspection of Figure 2 suggests that when measured ZTM findings for a given patient are plotted in this fashion, the pattern can be useful in the differential diagnosis of a given conductive hearing loss. Our clinical findings over the ensuing 40 years have reinforced this observation. Unfortunately, with the instrumentation available in the 1960s, measurements of this type, data reduction, and plotting were time intensive and rarely used in the clinic. Moreover, some instruments provided only one component of the complex acoustic impedance. Without both components of a complex number, it was impossible to speculate on the etiology of a conductive hearing loss for some patients. Also, it was impossible to convert the data from polar form to rectangular form. In 1972, the Grason-Stadler Co. introduced an electroacoustic instrument that measured complex acoustic admittance (YA) rather than acoustic impedance. This approach simplified numerical and (analog) electronic correction for the acoustic characteristics of the column of air interposed between the tip of the probe and the TM. In addition, YA data could be converted readily to the ZA form as long as both the acoustic susceptance (BA) and the acoustic conductance (GA) were reported.
Ten years later, as we began developing a standard for the American National Standard Institute (ANSI), it became clear that we needed a term that could be used to refer collectively to the array of acoustic measurements that were being made within the occluded, human EAM. The ANSI working group elected to borrow the term “immittance” from alternating-current electric-circuit theory. This term first was used by Bode (1945) as a convenient way to group im(pedance) and (ad)mittance measurements under one heading. It is important to recognize, however, that immittance does not have units because it applies to both impedance and admittance, which have different units. The resultant ANSI Standard was published in 1987 with the title: “American National Standard Specifications for Instruments to Measure Aural Acoustic Impedance and Admittance (Aural Acoustic Immittance)” ANSI (1987). In this supplement, we shall continue to use the term immittance when we refer to a variety of aural acoustic measurements under a single heading.
CHANGING AIR PRESSURE WITHIN THE EAM: TYMPANOMETRY
Visual inspection of the EAM and the TM has been used for the diagnosis of ear disease since the 14th century (de Chauliac 1363). Early otoscopes were simple conical specula. Today, optical otoscopes provide magnification of the field and use halogen light sources with fiberoptic technology. Video otoscopy allows the examiner to display his or her findings on a monitor and print the image. The otoscopic examination allows the clinician to inspect the external auditory canal and evaluate the condition of the TM. The color, the translucency, and the position (retracted, neutral, or bulging) of the drum membrane usually are reported.
The utility of the otoscope was enhanced when Siegel (1864) sealed the instrument hermetically, attached a rubber bulb and a soft aural speculum. With this pneumatic otoscope, air pressures above and below ambient could be introduced into the EAM. This innovation helped assess the mobility of the TM, the presence of a retracted TM, a tympanic perforation, and Eustachian tube function. It also aided the differential diagnosis of otitis media with effusion from acute otitis media.
Introduction of the pneumophone marked the first attempt to quantify some aspects of pneumatic otoscopy (van Dishoeck 1938; 1941). With this instrument, a pure-tone source and an air tube were mounted in a speculum that then was sealed hermetically into the EAM. When the air tube was connected to a u-tube manometer, the patient could use a finger to trace on the manometer the air pressure where the tone was loudest. At this point, the air pressure in the EAM was approximately the same as the air pressure in the patient’s middle ear.
Tympanometry marked a second attempt to quantify selected aspects of pneumatic otoscopy. With this technique, a pure-tone source and an air tube again are mounted in a speculum or probe that then is sealed hermetically into the EAM. A microphone also is mounted in the probe. The electrical output of this microphone is proportional to the SPL within the sealed EAM. Because changes in SPL are proportional to changes in the magnitude of acoustic impedance, plotting this electrical output while air pressure is changed yields a graph that is distinctive for conditions at the TM and within the middle ear. Some of the earliest tympanometric instruments actually plotted EAM SPL as a function of air pressure (Terkildsen & Thompson 1959).
Unlike the acoustic bridge (Zwislocki 1957) that was used for many early static acoustic-immittance measurements, the newer electroacoustic instruments designed for tympanometry achieved rapid acceptance in clinical audiology. There are several reasons for this acceptance. First, the instruments were relatively simple to use. Second, they could be used to generate tympanograms quickly, and, finally, they produced distinctive patterns for a variety of TM abnormalities, middle ear diseases, and conditions. This pattern recognition was attractive to clinicians.
Lidén was one of the first investigators to associate tympanometric patterns with specific middle ear conditions (Lidén 1969). He described pathognomonic patterns for four groups of adult, clinical subjects and identified these patterns with the letters A through D. A type E pattern was identified in subsequent publications from the same Swedish hospital (Lidén et al.1974). Figure 3 depicts the five types of tympanograms reported by Lidén et al. 1974. In the interim, Jerger attempted to replicate the findings of Lidén and his colleagues (Jerger 1970). He used a clinical electroacoustic immittance instrument that was in common use in the 1960s in the United States.
This instrument had a single probe-tone frequency of 220 Hz. His data supported the Swedish work for the type A (normal) pattern, for the flat type B (middle ear effusion and TM perforation) pattern, and for the type C (negative pressure) pattern but not for the type D and the type E patterns. This discrepancy was related to the frequency of the probe tones that were used. The 800 Hz probe tone that Lidén and his colleagues used was closer than a 220 Hz tone to the first resonance peak of the adult middle ear transmission system, and, thus, was able to produce consistently the W-shaped type D (ossicular disruption) pattern and the undulating type E (TM hypermobility) pattern.
Figure 4 depicts the single-component, low-frequency tympanograms that have been reported most commonly in the clinical literature (Jerger 1975). It also shows graphically how a tympanogram might progress from a type A (normal) pattern to a type B pattern during the development of serous otitis media. The abscissa on this graph is labeled “Compliance,” and it is not scaled. The data plotted probably are proportional to the magnitude of acoustic admittance. Type A, type B, and type C tympanograms are reported extensively in clinical reports, but today the abscissae usually are scaled in absolute physical units.
Until about 1972, many normal tympanograms were plotted with a single valley facing downward on the page. This convention, of course, was appropriate for measurement data that were proportional to ZA. In addition, almost all tympanograms were scaled in arbitrary units. In 1972, the Grason-Stadler Co. introduced an electroacoustic instrument that measured complex acoustic admittance (YA) in absolute units (acoustic siemen). Because YA is the reciprocal ZA, the normal, type A, tympanogram became a peak near ambient air pressure (in deka Pascal) rather than a valley. This instrument also had a second (678 Hz) probe tone that was three times higher in frequency than the standard 226 Hz tone. These changes led to several advancements. First, because tympanograms were plotted in absolute units and because a method was provided for correcting (approximately) for the admittance of the volume of air in the EAM, findings between clinics could be standardized. Second, inclusion of a probe tone that was closer to the first, middle ear resonance peak in adults allowed clinicians to produce type D and type E tympanograms. Third, the 678 Hz probe tone could be used to generate valid tympanograms with some neonates and young babies. In the 1970s, it became clear that the motility of the EAM in neonates often resulted in what appeared to be a normal low-frequency tympanogram even with fluid and embryonic mesenchyme in the middle ear (Paradise et al. 1976; Schwartz & Schwartz 1980; Sprague et al. 1985). Probe tones of 678 Hz and higher could be used to circumvent this problem.
A fourth advancement focused on quantification of tympanometric patterns. Using a Grason-Stadler instrument, investigators in Belgium plotted the rectangular components of YA, acoustic susceptance (BA), and acoustic conductance (GA), on the same graph for a given patient or subject and then counted the number of BA peaks and the number of GA peaks (Vanhuyse et al. 1975). With this technique and the 678 Hz probe tone, abnormal changes in mass and stiffness could be identified.
The next advancement in tympanometry did not force the examiner to worry about the frequency of the probe tone. Multifrequency tympanometry allowed the clinician to begin a test with the standard 226 Hz probe tone. The test could be terminated with only this low frequency if an adult patient without a conductive hearing loss generated a type A tympanogram. The test also could be terminated at 226 Hz for an adult or for a child older than about 5 months (Schwartz & Schwartz 1980) if a type B or a type C pattern was produced. For a patient with a conductive hearing loss and a type A tympanogram, the probe frequency then could be increased systematically until the first resonance peak of the middle ear was reached. An abnormal increase in the resonance frequency indicated a middle ear condition that increased the stiffness component of the system. This finding, in the presence of a type A, low-frequency tympanogram provided support for a diagnosis of clinical otosclerosis. An abnormal decrease in the resonance frequency indicated a middle ear condition that decreased the stiffness component of the system and provided support for a diagnosis of ossicular discontinuity (Colletti 1977; Funasaka et al. 1984; Lilly 1984; Hunter & Margolis 1992). The ability to increase the frequency of the probe tone also allowed clinicians to generate, in neonates, tympanograms that were not as affected by the excessive motility of the EAM at low frequencies. Moreover, the BA and GA patterns at each frequency permitted the clinician also to use the peak-counting analysis of Vanhuyse et al. (1975). In 1985, the Virtual Corporation developed a computer-based instrument that could be used for single-component tympanometry at 226 Hz, for multicomponent tympanometry at any frequency up to 2000 Hz, or to generate a family of tympanometric curves in 1/3 octave steps from 250 through 2000 Hz. Figure 5 shows a family of |YA| tympanograms from a subject with normal hearing. In panel A, the three-Dimensional plot has been sliced at 226 Hz to produce a normal, type A tympanogram. In panel B, the plot has been sliced just before the first resonance peak to produce a steeper tympanogram. Panel C shows the tympanogram at resonance. In panel D, the plot has been sliced at a frequency above resonance to produce a type D tympanogram.
In this brief introduction, we have attempted to review some of the major efforts that have been used to evaluate the condition of the human, adult middle ear transmission system, the middle ear cavity, and the function of the Eustachian tube. We have grouped most of this work under the rubric of “acoustic immittance.” To summarize:
- The earliest measurements of middle ear function were accomplished with ambient pressure in the EAM and with the middle ear muscles in a state of normal tonus. The results of these static acoustic-immittance measurements were useful in the development of closed acoustic cavities for the calibration of acoustic transducers. More recently, they have found clinical applications in the evaluation of probe-microphone measurements and hearing aids. When static, acoustic-immittance measurements were plotted, distinctive patterns were produced for normal middle ear systems and for patients with middle ear disease. Still, because of the time required for the measurements and for manual graphing of data, these plots never became commonplace in clinical reports.
- The introduction of air-pressure changes into the closed, adult EAM added a new and useful extension to static acoustic-immittance measurements. Single-component tympanograms obtained with low-frequency probe tones produced distinctive patterns in adults with normal middle ears (type A), with fluid-filled middle ears (type B), and with negative middle ear pressures (type C). Because these patterns could be generated rapidly and printed records easily could be made and archived, single-component tympanometry developed into a common clinical test.
- Multifrequency and multicomponent tympanometry added useful extensions to simple, single-component, low-frequency tympanometry. They permitted the clinician to use probe tones that were higher in frequency than 226 Hz. This addition to the test proved useful for neonates and for any patient who did not produce obvious low-frequency tympanograms. It also permitted the clinician to use the peak-counting analysis advanced by Vanhuyse et al. (1975) and to estimate the frequency of middle ear resonance. These advancements, in turn, helped with the differential diagnosis of middle ear disease.
- The measurement of wideband acoustic absorbance is not a totally new procedure. Rather, it is the latest enhancement to aural acoustic-immittance measurements. An enhancement that can expand our ability to characterize middle ear function and effects of ear disease on that function. It also allows us to evaluate middle ear function for frequencies whose wavelength is shorter than the length of the EAM. The articles that follow provide an elegant introduction to wideband, aural acoustic absorbance.
American National Standard Institute ANSI American National Standard Specifications for Instruments to Measure Aural Acoustic Impedance and Admittance (Aural Acoustic Immittance). (1987); S3:New York, NY American National Standards Institute 39–1987 (R-2012)
Bode H. W.. Network Analysis and Feedback Amplifier Design. (1945); Princeton, NJ Van Nostrand
Colletti V.. Multifrequency tympanometry. Audiol. (1977); 16:278–287
de Chauliac G.. A Text-Book of the Diseases of the Ear for Students and Practitioners. (1363); 5th ed. London, United Kingdom Baillière, Tindall & Cox
van Dishoeck H. A. E., van. Das Pneumophon. Ein Apparat zur Druckbestimmung im Mittelohr. Archive Ohren-Nasen-u Kehlhopft. (1938); 144:22–36
van Dishoeck H. A. E., van. Measurement of the tension of the tympanic membrane and the resistance of the Eustachian tube. Arch Otolaryngol. (1941); 34:354–357
Funasaka S., Funai H., Kumakawa K.. Sweep-frequency tympanometry: Its development and diagnostic value. Audiol. (1984); 23:366–379
Heaviside O.. C. Suskind. In: The Encyclopedia of Electronics (1962). (1886); Princeton, NJ Van Nostrand-Reinhold
Hunter L. L., Margolis R. H.. Multifrequency tympanometry: Current clinical applications. Am J Audiol. (1992); 1:33–43
Inglis A. H., Gray C. H. G., Jenkins R. T.. A voice and ear for telephone Measurements. Bell Syst Tec J. (1932); 11:293–317
Jerger J.. Clinical experience with impedance audiometry. Arch Otolaryngol. (1970); 92:311–324
Jerger J.. J Jerger. Diagnostic use of impedance measures. In: Clin Impedance Audiometry (pp. 156). (1975); New York, NY American Electromedics Corporation
Lidén G.. The scope and application of current audiometric tests. J Laryngol Otol. (1969); 83:507–520
Lidén G., Harford E., Hallén O.. Tympanometry for the diagnosis of ossicular disruption. Arch Otolaryngol. (1974); 99:23–29
Lilly D. J.. J. Jerger. Measurement of acoustic impedance at the tympanic membrane. In: Modern Developments in Audiology. (1973); 2nd ed. New York, NY Academic Press pp. 345–406
Lilly D. J.. Multiple frequency, multiple component tympanometry: New approaches to an old diagnostic problem. Ear Hear. (1984); 5:300–308
Paradise J. L., Smith C. G., Bluestone C. D.. Tympanometric detection of middle ear effusion in infants and young children. Pediatrics. (1976); 58:198–210
Schwartz D. M., Schwartz R. H.. Tympanometric findings in young infants with middle ear effusion: Some further observations. Int J Pediatr Otorhinolaryngol. (1980); 2:67–72
Siegel E. A Text-Book of the Diseases of the Ear for Students and Practitioners. (1864); 5th ed London, United Kingdom Baillière, Tindall & Cox
Sprague B. H., Wiley T. L., Goldstein R. Tympanometric and acoustic-reflex studies in neonates. J Speech Hear Res. (1985); 28:265–272
Steinmetz C.P.. C. Suskind. In: The Encyclopedia of Electronics. (1893); Princeton, NJ Van Nostrand-Reinhold
Terkildsen K., Thomson K.. The influence of pressure variations on the impedance of the human eardrum. J Laryngol Otol. (1959); 73:409–418
Vanhuyse V. J., Creten W. L, van Camp K. J.. On the W-notching of tympanograms. Scand Audiol. (1975); 4:45–50
Webster A. G.. Acoustical impedance and the theory of horns and of the phonograph. Proceedings of the National Academy of Science. (1919); 5:275–282
Zwislocki J. Some impedance measurements on normal and pathological ears. J Acoust Soc Am. (1957); 29:1312–1317