Published Date: 23 Mar 2015 Last Modified: 09 Jan 2018

CLINICAL DIAGNOSIS OF MIDDLE EAR DISORDERS USING WIDEBAND ENERGY REFLECTANCE

A Doctoral Thesis Presented to The Graduate College of Missouri State University In Partial Fulfillment Of the Requirements for the Degree

Copyright 2008 by [Alaaeldin Elsayed]

CLINICAL DIAGNOSIS OF MIDDLE EAR DISORDERS USING WIDEBAND ENERGY REFLECTANCE

Communication Sciences and Disorders

ABSTRACT

Accurate diagnosis of middle ear disorders in adults and children is a challenging task because of the complexity of disorders. Wideband energy reflectance (WBER) technique provides simplicity and accuracy in diagnosing middle ear disorders across wide frequency range. This research is expanding the studies of WBER to investigate the middle ear function in normal and pathological conditions of the middle ear in adults and children. Findings showed that WBER not only can distinguish abnormal from normal middle ear function but also can characterize different middle ear disorders in adults and children. Several specific WBER patterns were established in a variety of middle ear disorders among adults and children that will help in early diagnosis of such pathologies. The ER pattern was including significant higher ER in the children control group than the adult control group at 0.5 kHz and 1 kHz, abnormally high or shallower in otosclerotic ears, abnormally low in ears with TM perforation and abnormally low ER with deep notch in ears with hypermobile TM. In presence of negative middle-ear pressure, elevated ER at ambient pressure is also expected. Results also showed that standard tympanometry was less sensitive in diagnosing middle ear disorders when compared to WBER especially in otosclerotic cases. Further studies are still required to validate the clinical use of ER in larger number of individuals with confirmed middle ear disorders.

KEYWORDS: wideband energy reflectance, otosclerosis, otitis media with effusion, eustachian tube dysfunction, tympanometry.

This abstract is approved as to form and content

Wafaa Kaf, MD, MS, PhD

Chairperson, Advisory Committee

Missouri State Universit

CLINICAL DIAGNOSIS OF MIDDLE EAR DISORDERS USING WIDEBAND ENERGY REFLECTANCETITLE OF THESIS

By

Alaaeldin Elsayed

A Doctoral Thesis Submitted to the Graduate College Of Missouri State University In Partial Fulfillment of the Requirements For the Degree of Doctorate, Audiology

ACKNOWLEDGEMENTS

I would like to thank so many who encouraged me along this dissertation. First and foremost, I am thankful to God for all his blessings.

I am very grateful to Dr. Neil DiSarno for all his support and kind caring throughout my graduate school education.

Further, I am indeed grateful to Dr. Wafaa Kaf, my doctoral advisor, for her guidance, encouragement, and support throughout this work.

In addition, I would like to show appreciation to my committee members for their helpful comments and direction for this dissertation. Special thanks also to the faculty and secretarial staff of the Department of Communication Science and Disorders.

Thanks to Dr. Walid Albohy, and Dr. Ahmad Alhag for their help in collecting data for this study.

Special thanks and appreciation for my wife Enass and my children Mohamed and Nada, your love and delightful spirits has kept me going forward.

DEDICATION

This work is dedicated To My dear parents,

My beloved "Enass, Mohamed, and Nada",

Who made all of this possible,
for their endless encouragement and patience.

REVIEW OF THE LITERATURE

Hearing mechanism and the middle ear

Sound transmission. The hearing process includes the transmission of sound energy through the auditory canal to the tympanic membrane (TM). This sound energy results in vibration of the TM with an equal atmospheric pressure on both sides of the TM. The mechanical vibrations are, then, transmitted from the TM to the air-filled middle ear space and ossicles (malleus, incus and stapes), which further amplify the sound energy and transmit it, via oval window, to the fluid-filled inner ear. At the inner ear, the mechanical vibration is converted into electric waves and transmitted as nerve signals that are interpreted by the brain as sounds.

Mechanical properties of middle ear. The middle ear is an air-filled cavity that connects the outer ear canal to the labyrinth of the inner ear. This connection is established through the middle ear ossicels-malleus, incus and stapes. The malleus is attached to the TM by its handle; the incus bone lies in the middle between the malleus and the stapes while the footplate of the stapes is attached to the oval window of the inner ear. The middle ear cavity is also connected to the nasopharyngeal cavity through the Eustachian tube (Musiek and Baran, 2007). The Eustachian tube is important in maintaining an equal pressure on both sides of the TM and ventilation of the middle ear cavity. The tube also drain the middle ear into the nasopharynx (Channell, 2008). Figure 1 demonstrates schematic representation of the anatomy of the ear.

When the sound pressure moves the TM the mallus and incus consequently move together as one unit around a pivotal point. In doing so, both bones act as a lever; the lever arm formed by the manubrium of the malleus is slightly longer than that of the incus (about 1:1.3 ratio). In turn, the rotation of the long process of the incus around its pivotal point leads to the back and forth (piston-like) movement of the stapes footplate in the oval window of the inner ear. The movement of the stapes footplate is directly proportional to the frequency and amplitude of the sound waves. This route of sound transmission is called the "ossicular route". Acoustic route is another way of transmitting sound waves directly from the TM and the oval window to the cochlea. The direct acoustic stimulation of the oval and round windows, by passing the ossicles (acoustic route), plays a part in sound transmission In normal ears both routs are functioning but the upper hand is for the ossicular route (Voss, Rosowski, Merchant, and Peake, 2007).

From the above information, it appears that the middle ear plays important role in the hearing process. The middle ear mainly helps to correct the impedance mismatching between the air-filled middle ear and the fluid-filled cochlea and to transform the acoustic energy at the TM into mechanical energy that will eventually be transferred to the inner ear. The Impedance matching function of the middle ear is carried out by three mechanisms: the lever action of the ossicles of the middle ear, the area difference between the TM and the area of the stapes footplate, and the buckling of the curved TM. An outcome of these mechanisms is that the vibration obtained from the large area of the TM is focused to the much smaller oval window of the inner ear (21:1 area ratio), resulting in a differential pressure between the oval window connected to scala vestibuli and the round window connected to the scala tympani. This pressure differential is critical in maximizing the flow of sound energy and activation of the cochlear structures (Cummings, 2004). Accordingly, middle ear disorders are expected to affect the normal transmission of sound, resulting in conductive hearing loss (discussed below).

An illustration of the anatomical structure of External, Middle and Inner ear. Modified from "Medline Plus Medical Encyclopedia: Ear anatomy".

In addition to correcting the impedance mismatch between the air-filled middle ear and the fluid-filled cochlea, the middle ear also protects the inner ear from loud sound via the acoustic reflex. This mainly occurs as a result of reflex contraction of the two middle ear muscles, the tensor tympani and the stapedius, in response to loud sound leading to increased stiffness of the oscicular chain, and hence diminished sound transmission (Allen, Jeng, and Levitt, 2005). Given that the acoustic reflex mainly decreases the transmission of low frequency sounds thus, it improves speech discrimination in loud, low-frequency noisy environments. Unfortunately, the reflex does not protect the ear against impulsive sounds as gun shots due to prolonged latency in muscle contraction (Lynch, Peake, and Rosowski, 1994).

Pathophysiology of middle ear disorders

To further understand the pathology of middle ear disorders, it is important to consider the middle ear system as a vibrating mechanical system. Such a system is composed of three elements: mass, stiffness, and friction. When the mass and stiffness components are equal, so-called resonant frequency of the middle ear, it is expected that the amplitude of vibration of the middle ear is at maximum. On the other hand, when there is an increase in the mass without change in stiffness or friction the resonant frequency is lowered and the amplitude of vibration is lowered at frequencies above the resonant frequency. In contrast, when there is an increase in the stiffness component of the middle ear the resonant frequency increases and the magnitude of vibration reduces for frequencies below the resonant frequency (Roeser, Valente, and Hosford-Dunn, 2000).

Middle Ear Disorders are a variable group of pathological conditions that includes, for example, middle ear infection (Otitis Media with Effusion: OME), chronic otitis media with perforation of the TM, Eustachian Tube Dysfunction (ETD), ossicular disruption or dislocation and or/ otosclerosis. Such middle ear disorders may lead to conductive hearing loss due to their effects on mass, stiffness, and/or friction elements of the normal middle ear.

Perforated TM is induced by chronic otitis media or trauma to the ear. As a result, the normal structure and the function of the TM are altered. The degree of hearing loss is directly related to the size of the perforation (Voss et al., 2000) The perforation leads to equalization of pressure on both sides of the membrane which consequently leads to disturbance of the ossicular route and hearing loss (Voss et al., 2000). Normally the inward movement of the stapes is followed by an outward movement at the round window (push and pull mechanism). In the presence of TM perforation, this push and pull mechanism of the ossicles is disturbed and the sound waves energy reaching the oval window is reduced.

Ossicular dislocation usually follows a violent trauma to head or as a consequence of chronic otitis media and/or cholesteatoma. Disarticulation of the incudostapedial joint due to traffic accident was the most common pathlogy of ossicular disruption(Yetiser s, 2008). With the exception disruption due to chronic otitis media, the dislocation of the ossicles may or may not be accompanied by TM rupture. The injury results in loss of the impedance matching mechanism of the middle ear and a conductive hearing loss of about 40-60 dB (Merchant, Ravicz, and Rosowski, 1997).

Otosclerosis is a progressive disease of bone resorption and reformation that affects bones derived from the otic capsule. The etiology of the disease is not fully understood. The disease leads to osteodystrophy and fixation of the stapes in the oval window. Among the most accepted eatiological factors is genetic factors and viral infection. Otosclerosis is characterized clinically by progressive hearing loss, tinnitus and vertigo (Menger and Tange, 2003). Both conductive and sensory neural hearing loss has been reported in otosclerotic patients (Ramsay and Linthicum, 1994). Otosclerosis may affect the cochlea and other parts of the labyrinth as well (Menger and Tange, 2003). The resulting fixation of the footplate of the stapes leads to increased stiffness of the ossicular chain early in the disease. Increased stiffness of the middle ear affects the transmission of low frequency sounds. At later stages of the disease, the bone starts to grow adding a mass effect. This increase in mass of the middle ear affects the transmission of high frequency sounds as well (Shahnaz and Polka, 1997).

More disorders include inflammatory conditions of the middle ear such as otitis media (OM) and media with effusion (OME), chronic otitis media, and cholesteatoma. OM usually results from upper respiratory infections or allergies that lead to obstruction of the Eustachian tube (Channell, 2008). As a consequence, negative pressure develops in the middle ear resulting in otalgia due to stretching of the TM and mild hearing loss due to the increased stiffness of middle ear transmitting mechanism. If the negative pressure inside the middle ear is not relieved, a transudate accumulates inside the middle ear. The condition is then called OME. The hearing is further affected by the mass- friction effect. The degree of hearing loss depends on the type and the amount of the transudate. The combination of fluid and pressure in the middle ear was found to reduce TM movement at the umbo by 17 dB over the auditory frequency range (Dai, Wood, and Gan, 2008).

Middle ear function measures

Tuning fork testing. The tuning fork testing is one of the traditionally used qualitative hearing tests. They are used to examine the conductive component of hearing loss (external or middle ear pathology). Several tests have been descried including: Rinne, Schwabach, Bing, and Weber tests.

For Rinne test, the vibrating tuning fork is held against the skull, usually on the mastoid process bone behind the ear to cause vibrations through the bones of the skull and inner ear. To cause vibrations in the air next to the ear, the vibrating fork is then held next to, but not touching, the ear. In the test the patient is asked to determine if the sound heard through the bone is louder or that heard through the air. The results of the test are categorized as positive, negative, or equivocal. A negative Rinne test is indicated when the sound is heard louder by bone conduction than by air conduction which suggests a conductive component of the hearing loss. Although Rinne test was found to be highly specific in one study; the same author has suggested that it should be carried out only as a pack up test for pure tone audiometry in audiological evaluation of hearing loss (Browning and Swan, 1988; Thijs and Leffers, 1989). The Schwabach tuning fork test compares patient's bone conduction to the normal examiner. Bing tuning fork tests determines the presence or absence of the occlusion effect. Weber tunning fork test determines the type of a unilateral hearing loss. While Rinne test compares air conduction to bone conduction in the same patient.

Although the tuning fork testing is easy and reliable; it is still a subjective test that depends on the response of the patient and the degree of hearing loss. Additional drawbacks are that tuning fork testing is a qualitative and not a quantitative test, and does not diagnose the etiology of the conductive hearing loss.

Pure-tone Audiometry. Pure-tone Audiometry is a behavioral test that measures hearing threshold. The test has been used to diagnose type and degree of hearing loss for more than one hundred years.

During test setting, the patient is subjected to different tones to test the hearing mechanisms via air-conduction and bone conduction. Typically, the normal level of pure tone audiogram air and bone conduction will lie between 0-15 dB HL for children and 0-25 dB HL for adults. According to Northern and Downs (1991), the degree of hearing loss can be classified in adults as (0-25 dB HL) within normal limits, Mild (26-40 dB HL), Moderate (41- 55 dB HL), Moderate-Severe (56-70), Severe (71-90 dB HL) or Profound (91 + dB HL) hearing loss. In children it is classified as normal (0-15 dB HL), Slight (15-25 dB HL), Mild (25-30 dB HL), Moderate (30-50 dB HL), Severe (50-70 dB HL), Profound (70 + dB HL) hearing loss. This classification is applied to PTA of 500, 1000, and 2000 Hz (Roeser et al, 2000).

Different types of hearing loss are interpreted by comparing air conduction thresholds to bone conduction thresholds. When the air conduction threshold elevated to a maximum around 60-70 dB HL in the presence of normal bone conduction threshold, this type of hearing loss is called conductive hearing loss. In sensorineural hearing loss the pure tone audiogram shows both air and bone conduction thresholds are elevated and with a 10 dB HL or less in between. Mixed hearing loss displays elevation in both air and bone conduction thresholds, but with the bone conduction threshold at better intensities than the air conduction by 10 dB HL or more. In both conductive and mixed hearing loss, the difference in air and bone conduction thresholds is called air-bone gap; and it represents the amount of conductive hearing loss present (Roeser et al, 2000).

The use of pure-tone audiometry provides quantitative information regarding the degree and type of hearing loss. However, it does not diagnose the cause of hearing loss and cannot be used in infants, young children, and difficult-to-test subject. Mannina (1997) reported that the diagnosis of middle ear disorders in school-aged children is less efficient when using pure-tone audiometry alone. To improve the diagnosis of middle ear disorder, Yockel (2001) demonstrated that the addition of tympanometry to audiometry does improve the diagnosis of OME than using audiometry alone.

Assessing Middle ear function is a very important step in early diagnosis and treatment of conductive hearing loss. Since the usually used subjective tests, the tuning-fork and pure tone audiometry, cannot identify the etiology of underlying middle ear disease, other objective measures such as acoustic immittance are needed for differential diagnosis and accurate diagnosis of specific middle ear disorders.

Acoustic Immittance. Several objective measurements of middle ear function have been developed over the last four decades. Various anatomical structures of the middle ear represent complex network system that affects the sound presented to the ear. Not all the sound represented to the middle ear is delivered to the cochlea, but some of the power is absorbed by the bony structure of the middle ear (Zwislocki, 1982). Acoustic Immittance using tympanometry assess the middle ear status by measuring the transmitted sound energy to the middle ear.

Acoustic Immittance provides objective information about the mechanical transfer function in the outer and middle ear. Acoustic Immittance is defined, as the velocity with which an objects moves in proportional to an applied force, while Acoustic Impedance (Za) is the opposition offered by middle ear and the TM to the flow of energy. Mathematically acoustic admittance (Ya) of a system is the reciprocal of impedance. Acoustic Immittance refers collectively to acoustic admittance, acoustic impedance or both ("Tympanometry. ASHA Working Group on Aural Acoustic-Immittance Measurements Committee on Audiologic Evaluation", 1988). Investigators have found that abnormalities in the middle ear transmission might be reflected in the acoustic condition of the TM (Allen et al, 2005). Acoustic Immittance can be measured to single probe-tone frequency (single frequency tympanometry) or to series of multiple probe frequencies (multifrequency tympanometry).

Single frequency tympanometry. Tympanometry is one of the earliest objective methods used to evaluate middle ear function. Tympanometry measures the acoustic immittance of the middle ear as a function of changing the air pressure in the ear canal. A single probe tone tympanometry is the conventional measure of middle ear function in response to low frequency probe tone, 226 Hz, under varying static air pressure. Evaluation of the acoustic immittance of normal and different middle ear disorders was done by Otto Metz, 1946, and confirmed later by Feldman, 1963 (Katz, 2009)

In 1970, James Jerger began to incorporate immittance measurement into the routine audiological evaluation. Jerger classified tympanograms as type A, B, or C depending on the shape of the tympanogram (with or without peak) and location of the peak when present. Type A is the normal tympanogram with the peak at or near the atmospheric pressure (+25 to -100 daPa). Type A is further divided into subtypes Ad and As for high and low peaked type A tympanograms respectively (Feldman, 1976). Type B tympanogram has no peak and relates to middle ear effusion, infection with normal ear canal volume, or due to large TM perforation with large ear canal volume. Type C is a negatively shifted tympanogram that reflects Eustachian tube dysfunction, a precursor of serous OM, mostly evolved from type B (Katz, 2009).

Since 1970, single frequency Tympanometry is the conventional clinical middle ear measure because it is a non-invasive, objective, and cheap indicator of many middle ear pathologies in children and adults. Unfortunately, low frequency probe tone tympanometry has high false negatives in infants younger than seven months (Holte, Margolis, and Cavanaugh, 1991). This is explained by the movement of the infant's ear canal wall with pressure changes in the external ear canal due to immaturity of the bony part of the external auditory canal. In addition, tympanometry was found to be relatively insensitive to many lesions that affect the ossicular chain of the middle ear (Lilly, 1984). Furthermore, Keefe and Levi (1996) reported false positive tympanometry results compared to energy reflectance, a recent middle ear function measure. They found normal middle ear energy reflectance at higher frequencies in infants with flat low probe tone tympanometry.

Multifrequency tympanometry. Multifrequency Tympanometry (MFT), which was first introduced by Colletti in 1976, measures middle ear impedance using multiple frequency probe tones ranging from 226-Hz to 500 Hz and up to 2000 Hz (Colletti,1976) . Similar to previous discussion about the three elements of the mechanical system of the middle ear, admittance of the middle ear has three components: stiffness (compliant susceptance), mass susceptance and conductance (resistance).

A tympanometric pattern was developed by Vanhuyse and colleagues in 1975 that helped in interpreting the underlying middle ear pathology using MFT. The Vanhuyse tympanometric pattern is based on the assumption of the shapes and locations of reactance (X) and resistance (R) tympanograms. Using a conversion equation the model can predict the shapes of susceptance (B) and conductance (G) tympanograms. Vanhuyse et al proposed four normal patterns: 1B1G, 3B1G, 3B3G, and 5B3G as shown in Figure 2. 1B1G pattern is the normal tympanogram with a one susceptance (B) and one conductance (G) peak. It occurs when reactance (X) is negative and its absolute value is greater than resistance (R) at all pressure used (the ear stiffness is controlled). As the probe frequency increases the curve becomes more complex and notched. 3BIG model has three peaks of susceptance (B) and one conductance (G) peak. It represent negative reactance (X) with an absolute value greater than resistance (R) at low pressure and smaller than resistance (R) at high pressure. The third model (3B3G) appears when the ear is mass-controlled. In 3B3G model the reactance is positive and less than resistance (X < R) at low pressure and negative at high pressure. 5B3G pattern occurs when the reactance is positive and greater than resistance (X > R) at low pressure and becoming negative at high pressure (Margolis, Saly, and Keefe, 1999). Figure 2.

A graphic presentation of the model presented by Vanhuyse, Creten and Van Camp (1975). The resistance (R) , negative resistance (-R) and the reactance (X) tympanograms is shown in the upper left corner of each panel. Negative R is shown to compare the magnitude of the reactance X. The corresponding admittance (Y), (lower left corner), susceptance (B), (upper right corner) and conductance (G), (lower right corner) are also shown in each panel. Four patterns are presented and classified according to the number of extrema in the susceptance B and conductance G tympanograms. The pattern (1B1G) in panel one shows both susceptance and conductance have single extrema and reactance is negative. The pattern (3B1G) in panel two shows conductance G is single peaked with three extrema in susceptance B, reactance X is still negative but its absolute value is greater than resistance at high pressure. The pattern (3B3G) in panel three shows three extrema in susceptance B, conductance G, and admittance Y tympanograms, reactance Y is positive but less than resistance R . The pattern (5B3G) in panel four shows five extrema in susceptance B tympanogram and three extrema in conductance G, and admittance Y tympanograms, reactance Y is positive and greater than resistance R at low pressure.

Because of the use of measuring middle ear function to several probe tone frequency, MFT is considered superior to single frequency tympanometry in detecting high impedance pathological conditions of the middle ear such as middle ear effusion, otosclerosis, and cholesteatoma. Such pathological conditions were not detected by conventional tympanometry (Colletti, 1976, Keefe and Levi, 1996, Shahnaz et al 2009). Several studies have shown that MFT has higher sensitivity and specificity in detecting middle ear pathologies such as TM mass or adhesions (Margolis, Schachern, and Fulton, 1998). Also, MFT is more sensitive than single frequency tympanometry in identifying normal and abnormal middle ear condition in neonates (Shahnaz, Miranda, and Polka, 2008). However, MFT is of limited clinical use for several reasons: long testing time, limited frequency range, and unreliable data above 1000 Hz (Allen et al, 2005). The use of wideband energy reflectance is shown to address the above limitations of MFT.

Wideband energy reflectance. The wideband energy reflectance (WBER) is a new technique that has been introduced recently to evaluate middle ear dysfunction (Keefe, Ling, and Bulen, 1992). Simply the idea of WBER is that incident sound to the ear is transmitted through the ear canal and TM, some of this sound energy is absorbed through the middle ear and cochlea and part of it is reflected back (Figure 3). The energy reflectance (ER) is defined as the square magnitude of pressure reflectance ¦R(f) ¦2, which represents the ratio of the sound energy reflected from the TM to the incident sound energy at frequency (f). ER ratio ranges from one to zero (1.0 = all incident sound energy is reflected, and 0.0 = all sound energy is absorbed) (Allen et al, 2005). ER is an indicator of the middle ear power to transfer sound (Feeney, Grant, and Marryott, 2003).

Energy reflectance (ER) measurers middle ear function over a wide band of frequencies (0.2- 8 kHz). ER is the ratio of the reflected energy (red arrow) to the incident energy (yellow arrow). When all incident sound energy is reflected back ER ratio equals one. When all incident sound energy is absorbed ER equals zero. Red arrow represents reflected sound energy; yellow arrow represents incident sound energy; green arrow represent absorbed sound energy. Modified from "Medline Plus Medical Encyclopedia: Ear anatomy".

WBER measures middle ear function using a chirp stimulus at 65 dB SPL over a wide frequency range, typically 0.2 to 8 kHz and at fixed ambient pressure (Feeney et al, 2003) . Normative data has shown that most incident acoustic power is reflected back to the ear canal (ER ratio closes to 1) at frequency range below 1 kHz or above 10 kHz that also show poor hearing threshold or at frequencies below 1 kHz and above 4 kHz (less efficient middle ear function) (Keefe, Bulen, Arehart, and Burns, 1993). More specifically, 50% of the acoustic power is transmitted to the middle ear between 1-5 kHz frequency range, indicating that the most effective middle ear transfer function (ER is at its lowest values, closer to one) occurs around 1-5 kHz (Allen et al, 2005; Keefe et al, 1993; Schairer, Ellison, Fitzpatrick, and Keefe, 2007).

WBER has been used in measuring normal middle ear function and middle ear disorders using ambient pressure (Allen et al, 2005; Feeney et al, 2003; Shahnaz et al., 2009). In other studies the researchers used pressure to measure the acoustic stapedial reflex (Feeney and Sanford, 2005; Schairer et al, 2007). Development of the middle ear in infants was also investigated using WBER (Keefe and Abdala, 2007; Keefe e al, 1993; Keefe and Levi, 1996).

Wideband energy reflectance in neonatal screening

Keefe et al. (1993) and Keefe and Levi (1996) reported that the acoustic response properties of the external and middle ear varies significantly over the first 2 years of life. These changes, mostly physical changes, are responsible for the mass-dominant infant's middle ear system with lower resonant frequency. The main components of this mass-dominant effect is the pars flaccida of the TM, ossicles, and perilymph in the cochlea (Van Camp, Margolis, Wilson, Creten, and Shanks, 1986). The mesenchyme in infant's middle ear may add to the mass effect (Meyer, Jardine, and Deverson, 1997). This is completely in contrast to adult's middle ear, which is a stiffness-dominant system at low frequency (Holte et al, 1991; Keefe and Levi, 1996). The TM, tendons and ligaments, the space between the mastoid and the middle ear cavity, and the viscosity of the perilymph and the mucous lining of the middle ear cavity constitute the stiffness component of the middle ear (Van Camp, Margolis, Wilson, Creten, and Shanks, 1986).

Recently, Shahnaz (2008) have compared MFT and WBER findings between normal adults and normal-hearing neonates in the neonatal intensive care units (NICU), who passed the neonatal hearing screening test. The researcher found maximum absorption of the incident energy at narrower range of frequencies (1.2 - 2.7 kHz) in normal babies compared to adults (2.8 - 4.8 kHz) (Shahnaz, 2008; Shahnaz et al, 2008). This preliminary normative data from 49 neonatal ears reflects the potential diagnostic benefits of the WBER test in detecting middle ear effusion in neonates.

Wideband energy reflectance in otosclerosis

Although the main definitive diagnosis of Otosclerosis is during surgery, an accurate preoperative audiological diagnosis is very important indication for surgery. Still, pure-tone audiometry has its own limitations that prevent accurate diagnosis of otosclerosis. Also, standard 226 Hz tympanometry is usually within normal type A tympanogram in most otosclerotic patients (Jerger, Anthony, Jerger, and Mauldin, 1974). While multiple frequency tempanometry may be helpful in diagnosing otosclerosis, it adds little information to the diagnosis (Probst, 2007). On the other hand, the WBER responses in three ears of otosclerosis fell outside the 5th to 95th percentile of the normative data and presented a distinctive pattern for the disease (Feeney et al, 2003); which suggests that WEBR is a sensitive middle ear measure. In a recent study WBER was found to be helpful in distinguishing 28 otosclerotic ears from normal and/or other causes of conductive hearing loss. A significantly higher ER was found in otosclerotic ears at frequency range of 0.4- 1 kHz as compared to normal ears. In the same study WBER was found to be more sensitive in diagnosing otosclerosis than the conventional 226 Hz tympanometry and the MFT (Shahnaz et al., 2009).

Wideband energy reflectance in other middle ear pathology

Hunter and colleagues (2008) found higher sensitivity of WBER in detecting otitis media in infants and children with cleft palate (Hunter, Bagger-Sjoback, and Lundberg, 2008). Feeney and colleagues in 2003 studied WBER at ambient pressure in 13 ears with different middle ear disorders and comparative normal. Different middle ear disorders were involved in this study included: four ears with OME, one ear with ossicular discontinuity, two ears with otosclerosis, two ears with hypermobile TM, two ears with perforated TM, and one participant with bilateral sensorineural hearing loss. The results suggested a distinctive WBER pattern in different pathologies of the middle ear (Feeney et at, 2003).

There are several advantages of expressing measuring the ER of middle ear function using WBER measure: (1) WBER is not affected by external auditory canal properties, also not sensitive to probe location in the external auditory canal; (2) WBER is more closely related to hearing sensitivity than acoustic immittance; (3) In contrast to tympanometry, WBER can be carried out at ambient (atmospheric) pressure, thus, it can be used in young infants because it is affected by the floppy ear canal wall; and (4) ER is not affected by the probe position and standing waves in the outer ear canal, whereas admittance or impedance measurements is affected by these variables (Allen et al, 2005; Margolis et al, 1999) .

In summary, several Audiological assessment methods have been used to evaluate the middle ear disorders. Some of the tests that have been used are subjective tests like the tunning fork and pure tone audiometry. Researchers have depended on these two methods until recently to differentiate conductive from sensorineural hearing loss. The problem with these two traditional tests is that they depend on the response of the patient and that the underlying pathology of hearing loss can not be identified.

The recent advancement in otological surgery raised the importance of specific diagnosis of the underlying middle ear pathology. Tympanometry was a new development in the field of audiological assessment of the middle ear disorders. The test is an objective one that dependence on measuring the theory of acoustic impedance. Single low frequency tympanometry carried out at 226 Hz can only identify low frequency middle ear pathology, but fails to accurately diagnose middle ear disorders in neonates as well as high frequency middle ear pathologies. Multifrequancy tympanometry was a new addition to the Middle ear test battery. MFT was shown to be very helpful in detecting high frequency pathologies of the middle ear. However, MFT is still invalid in neonatal screening and in several middle ear disorders as otosclerosis and ossicular discontinuity.

Wideband energy reflectance is a potential advancement in the field of middle ear disorder audiological diagnosis. The WBER method depends on the new aspect of sound energy reflectance. WBER is a promising technique that will help in diagnosing middle ear dysfunctions in adults, children, and infants. Most research on WBER has concentrated on normative data and neonatal studies. The data available about the profile of WBER in different middle ear pathologies is scars. The current study is carried out with the objective of expanding the investigation of WBER as a diagnostic method in Middle ear disorders in children and adults.

Study objectives

WBER is potentially useful in differentiating normal from middle ear dysfunction in adults and infants. The diagnostic role of WBER in different middle ear pathologies has not been extensively studied. The current study is carried out with the objectives of: 1) Establishing the WBER patterns in a variety of middle ear disorders that will help in early diagnosis of a specific type of middle ear disorder; 2) Comparing the sensitivity and specificity of WBER to that of standard tympanometry using 226 Hz and 1000 Hz probe tone frequency; and 3) Providing normal pattern of WBER in control groups (adults and children).

MATERIALS AND METHODS

Participants

Given that most middle ear disorders in adults, such as tympanic membrane perforation and otosclerosis-race dependant and prevalent in Egypt and the Middle East-are not common disorders in the U.S; therefore, it deemed important to recruit those patients from Egypt and to run the test protocol at the ear clinic at the Zagazig University Hospitals. Study proposal and consent form were approved from the Institutional Review Board (IRB) at both Missouri State University (Date; approval # ) and the Ear, Nose, and Throat (ENT) department at Zagazig University Hospitals, Zagazig, Egypt. Two experimental groups-adults and children-and their matched control group, were recruited to participate in the study. The adult experimental group consisted of patients with clinically confirmed middle ear disorders by the ear doctor at the ENT outpatient clinic at the Zagazig University Hospitals. The participants for the adult control group were recruited from friends, relatives, and colleagues through verbal announcements. Only Egyptian participants were included in the adult control group to have homogenous group since the experimental adult participants were Egyptian. The participants for the children group, experimental and control were recruited from day care centers in Springfield, MO. Signed consent form was obtained from adult participants and children's parents before starting the test protocol of the study.

Inclusion criteria. Two main groups of adult volunteers were recruited to participate in this study. The experimental adult group consisted of 22 participants (10 males, 22 females; 32 ears were tested), whose age ranged from 20 to 57 years (Mean 36.7 ± 11.7 SD). This group was further subdivided into subgroups based on their middle ear disorder: otosclerosis, ETD, hypermobile TM, and TM perforation. Table 1 shows the participant characteristics of each group. Another age matched adult control group (N = 13 male ears, 4 female ears) was included in the study, ages from 15-62 years (Mean 35.2 ± 10.8 SD). All control groups had a clear otoscopic examination with no gross eardrum abnormalities or excessive cerumen; negative history of head trauma, hearing loss or middle ear diseases; normal hearing sensitivity = 25 dB HL at octave frequencies between 0.25- 8 kHz and no air bone gap more than 10 dB HL between 0.25-4 kHz; normal 226-Hz tympanometry (Margolis et al, 1999).

Table 1 also shows the characteristics for the children experimental group with a total of 17 ears (11 males, 6 females), ages from 1 to 4 years old (Mean 3.1 ± 1.2 SD) with a variety of middle ear disorders (OME, TM perforation, and ETD) as shown in Table 1. Another age and gender matched control group (N = 9 male ears, 12 female ears) was included in the study, ages from 1-5 years (Mean 3.8 ±1 SD). All control groups had a clear otoscopic examination with no gross eardrum abnormalities or excessive cerumen; negative history of head trauma, hearing loss or middle ear diseases; normal hearing sensitivity = 25 dB HL at octave frequencies between 0.5- 4 kHz; as the test wasn't done in sound-treated room; normal 226-Hz tympanometry.

All participants underwent the following procedures:

Otoscopic examination of the ear canal and TM

A handheld light source (Otoscope) was used to examine ear canals and ear drums to confirm the normal or the clinical diagnosis of a particular middle ear disorder. The physician examined the color, position of the tympanic membrane, as well as the presence of any apparent middle ear pathology. Experimental groups with combined pathologies (e.x., otosclerosis and TM perforation) were excluded from the study since tympanometry is sensitive to the most lateral pathology therefore, the lateral pathology can overcome the medial pathology (Feldman, 1974). Individuals with impacted wax in the ear canals were referred for medical attention to remove the wax before conducting any additional testing. Participants in the otosclerotic group were diagnosed clinically according to the accepted clinical criteria (Menger and Tange, 2003).

Standard tympanometry

GSI-Tympstar Tympanometer: 226 Hz tympanometry with a high frequency option was used to evaluate middle ear status in children. The Tympstar was calibrated according to the American National Standards Institute (ANSI, 1987) standards at the atmospheric pressure to verify probe tone frequencies and intensity before data collection. Immittance testing was documented at standard 226 Hz tympanogram by placing a probe-tip in the participant's ear canal, and air pressure was varied slightly from +200 to -400 daPa to evaluate the middle ear function. Participants were required to sit quietly during the procedure. For judgments of the shape of tympanograms, Jerger classification (1970) was used (type A, B, C). Type A tympanogram has a normal peak height and location on the pressure axis, type B tympanogram is flat, and type C tympanogram has a peak towards negative pressure. Also we used Feldman's (1976) subtypes Ad and As for abnormally high- and low-peaked type A tympanograms. Our adult norms for static admittance were classified according to Margolis and Goycoolea, (1993) (0.30-1.70 mmho), while the admittance criteria for the children groups were classified according to Roeser et al, (2000) (0.25-1.05 mmho). The norms data for equivalent ear canal volume were 0.9-2.0 cm³ for adults (Margolis and Heller, 1987) and 0.3-0.9 cm³ for children (Shanks, Stelmachowicz, Beauchaine, & Schulte, 1992).

Pure tone audiometry

Pure tone Audiometry (Standard Hearing Test) was conducted using a Madsen Grason Stadler GSI-61 audiometer calibrated according to ANSI standards (re: S3.6.1989). The Pure tone Audiometry was carried out in single-walled booth using TDH-39 headphones. The procedure involves responding to the presence of auditory stimuli by pressing a button every time the individual hears the stimulus. The procedure was conducted to measure the hearing sensitivity at different pitches (0.25-8 kHz) for adult groups, (0.5-4 kHz) for children group, and at different loudness levels. Following Northern and Downs (1991), classification of hearing threshold was used for all groups as follows: 0-25 dB HL- within normal limits, 26-40 dB HL- Mild loss, 41-55 dB HL- Moderate loss, 56-70 dB HL- Moderate to severe loss, 71-90 dB HL- Severe loss, and 91 + dB HL- Profound loss.

Wideband energy reflectance

This is another measure of middle ear function. WBER is a computerized system for research that is used to measure wideband evoked aural responses in the ear canal. Two WBER systems were used: Reflwin version 1.9 was used to measure middle ear function in the children groups while Reflwin version 2.0 from Interacoustics was used for the adult groups. Both systems are composed of a desktop computer with data acquisition sound card. However, version 1.9 has Etymotic Research Probe system (ER-10C) with a 20 dB Receiver gain while the version 2.0 has a different probe system (see below). Recording was conducted in a sound-booth (version 1.9) or a quiet room (version 2.0). Calibration was done daily before testing (see below).

WBER hardware and software configuration

The WBER Instrument includes the following components (Figure4):

1. A CardDeluxe professional sound card is driving the probe and recording the microphone output from the probe. CardDeluxe sound card have 2 channels in and out, with 22.05 kHz sample rate, and 24 bit voltage resolution.

WBER Research System. Panel (A) show the WB Interacoustics AT235 device modified for computer control (Tympanometry/ pump controller). Panel (B) show the Windows PC which includes the card sound and the in and out channels.

1. Interacoustics probe: 2 receiver ports, 1 microphone port, and static pressure coupled to port.
2. A dedicated PC with Monitor holds the CardDeluxe sound card installed and a serial port to communicate with the AT235 middle ear analyzer. The PC runs the ReflWin software under the windows XP operating system.
3. Interacoustic tympanometry pump/controller (version 2.0): An interacoustic AT235 Middle Ear Analyzer has been modified to allow control by ReflWin software for wideband tympanometry (WBT) measurements in the ear. A new dedicated probe is designed that allows for Wideband Absorbance (WBA) measurement, as well as WBT, which is a pressurized WBA test.
4. ReflWin holds Matlab code as a utility library of analysis, plotting and database routines, which allow the user to create electronic (PDF) documents with measurement results and create a computerized database readable by commercial spreadsheet software, i.e. Excel.

WBER calibration.WBER is initially calibrated at the factory. The machine was calibrated every day before testing procedure. Two sets of calibration tubes were used. Each set including a long and a short tube. One tube set has a large diameter (diameter of 0.794 cm) appropriate for testing adult-sized ear canals; the second set is a small diameter set of tubes (diameter of 0.476 cm) appropriate for testing infant-sized ear canals. Each set of large or small tubes has a long and a short tube that are closed at one end. The calibration was started by inserting the probe tip with suitable ear tip attached to it into the large long tube and then the large short tube. If the recordings were satisfactory, it can be saved to a file. Re-recording was done if the results were unsatisfactory. The same calibration steps were repeated for the small tubes.

WBER testing procedure. The probe tip with a suitable-size foam eartipe (version 1.9) or reusable plastic eartip (version 2.0) was securely placed in the participant's ear canal. Chirp sound was presented at moderate stimulus level (65 dB SPL) under ambient pressure. The patient was instructed to sit quietly during the test, which takes less than 2 minutes of testing per ear. Averaged response was obtained to 32 sweeps and measured to 6th octave frequencies. The probe was reinserted if a leak was observed according to the model suggested by Keef et al (2000). The energy reflectance (ER) ratio was measured over a wideband of frequency ranging from 0.2 to 8 kHz generated by the probe receiver. Results were recorded and automatically saved on the provided software. ER results ranged from 0.0 to 1.0; high reflectance values close to 1.0 indicate low absorbance by the middle ear and that most of the sound energy is reflected back, and vice versa.

Study design and statistical analysis

Descriptive statistics for WBER include calculation of the mean ER and the 10th- 90th percentile of the ER over the range of 0.2 - 8 kHz for both control groups (children and adults). The mean ER data for the experimental groups was calculated and compared to the matched control group to determine whether the patient's mean ER data was deviated from the normal 10th to the 90th percentile data. For inferential statistics, one-way repeated measure Analysis of Variance (ANOVA) was computed between each of the experimental groups (adults or children) and the matched control group (adults or children). Another comparison was carried out between the two control groups (adults versus children) using another one-way ANOVA. In both analyses, ER ratios were compared between the groups at each frequency, and the significance level was determined at alpha level of <0.05.

RESULTS

Adult Subjects

The data for the WBER were analyzed using several one way ANOVA to compare between the control groups (adults vs. children) and between the control groups and the corresponding pathological groups (otosclerosis, TM perforation, hypermobile TM, ETD) as a function of octave frequency 0.25, 0.5, 1, 2, 4 and 8 kHz.

Control groups (adults and children). Tympanometry results revealed that all adults group had normal type A tympanogram with normal ear canal volume (Yec) values (0.9 to 1.8 mmho) and admittance at the TM level (Ytm) values (0.4 to 1.2 mmho). Similarly, all children group had normal type A tympanogram with the Yec values ranged from 0.4 to 0.9 mmho, and the Ytm values ranged from 0.3 to 1 mmho. Pure tone thresholds were < 25 dB HL for all participants.

Energy reflectance (ER) patterns were similar between the two control groups as shown in Figure 5. The mean ER (50th %) was high in the low frequencies, decreased as a function of frequency to a minimum (the most absorbed energy) between 2 kHz to 3.8 kHz, and then increased gradually at higher frequencies. The main difference between the two control groups was the presence of the lowest ER between 1.5-3.5 kHz in normal adults as compared to 2 - 3.8 kHz in normal children (Figure 5 and Table 2). The ANOVA analysis shows a statistically significant difference between the two control groups only at 0.5 kHz (F [1, 36]) = 16.973, p = 0.000) and 1 kHz (F [1, 36]) = 5.489, p = 0.025. This difference is mainly due to higher ER for the children group (0.908) than the adult group (0.792) at 0.5 kHz. Also at 1 kHz, the ER was higher (0.557) than the adult group (0.448) as shown in Table 2.

Energy reflectance in normal adults and normal children. The lines represent the 10 th, 25 th, 50 th, 75 th, and 90 th percentile of the energy reflectance of the control adults (left panel) and control children (right panel).

Control adults vs. otosclerotic patients. Compared to the normal tympanometric findings from the control adults, all otosclerotic ears had notmal type A tympanogram. Their Yec volumes were (0.6 to 1.1 mmho), and Ytm magnitudes were (0.2 to 1.0 mmho). The average pure-tone hearing thresholds were < 25 dB HL for the control group, and 51 to 60 dB HL for the otosclerotic group. The average air bone gaps for the otosclerotic group were 28.5 to 46 dB. The mean ER for the otosclerosis group (n=10) was shallower with restricted minimum ER around 2.8 kHz compared to the control adults (Figure 6). Although the ANOVA results reveals that there was no statistically significant difference between the two groups, the mean ER (50 th percentile) were higher at frequencies 0.5, 1, 2, and 4 kHz (Table 3).

Energy reflectance in normal adults and adults with otosclerosis. The lines represent the 10 th, 25 th, 50 th, 75 th, and 90 th percentile of the energy reflectance of the control adults (left panel) and adults with otosclerosis (right panel).

Control adults vs. patients with perforated TM. Compared to the normal type A tympanogram of the control adult group, individuals with TM perforation had type B flat tympanogram with abnormally high Yec (3.4 to 5.9 mmho). Air conduction hearing thresholds ranged between 15 to 80 dB HL for the perforated TM group compared to < 25 dB HL for the control group. A lower ER ratio was observed in patients with TM perforation (n = 7) than that of the adult control group (n = 17). The means ER was low in the lower frequencies, increased as a function of frequency to a maximum at 1.8 kHz, and then decreased gradually to a maximum at 3.5 kHz, and then increased again at higher frequencies (Figure 7). The ANOVA results show statistically significant differences between the two groups at all frequencies except at 8 kHz (F [1,22] = 2.837, p = 0.106). This significant difference is mainly because the control group has higher ER than the TM perforated group at 0.25 kHz (0.993 vs. 0.577) and 0.5 kHz (0.792 vs. 0.566). The reverse pattern occurred at higher frequencies wherein the TM perforated group (0.717) had higher ER than the control group at 1 kHz (0.717 vs. 0.448), 2 kHz (0.603 vs. 0.330), and 4 kHz (0.719 vs. 0.453) (Figure 7 and Table 4).

The lines represent the 10 th, 25 th, 50 th, 75 th, and 90 th percentile of the energy reflectance of the control adults (left panel) and adults with perforated TM (right panel).

Control adults vs. patients with hypermobile TM. All control adults group have a normal type A tympanogram with Yec were (0.9 to 1.8 cm³), and Ytm were (0.4 to 1.2 mmho), while the hypermobile TM group had abnormal type Ad tympanogram with abnormally high admittance (Ytm: 1.8 -1.9 mmho) and normal Yec volumes (0.7 to 0.9 mmho). Pure tone thresholds ranged between 5 to 30 dB HL for the hypermobile TM group compared to the thresholds for the control group (< 25 dB HL). As shown in Figure 8 and compared to normal ER, the mean ER for the hypermobile TM group was abnormally low in the low frequencies up to around 2.8 kHz before it increased to a maximum above 6 kHz. Also, the figure shows that the ER is more variable in the low frequency range in the hypermobile group than the control group.

The ANOVA results between the control adult group (n= 17) and the hypermobile TM group (n= 7) show statistically significant differences between the two groups at 0.25 kHz (F [1, 22]) = 93.528, p = 0.000), 0.5 kHz (f [1, 22]) = 96.176, p = 0.000), and at 1 kHz (F [1, 22]) = 13.255, p = 0.001). The ER was not statistically different between groups at frequencies higher than 1 kHz. Table 5 and Figure 8 show that the differences between the two groups are mainly due to higher ER for the control group than the hypermobile group at 0.25 kHz (0.993 vs. 0.420), 0.5kHz (0.792 vs. 0.273), and 1 kHz (0.447 vs. 0.203).

The lines represent the 10 th, 25 th, 50 th, 75 th, and 90 th percentile of the energy reflectance of the control adults (left panel) and adults with hypermobile TM (right panel).

Control adults vs. patients with ETD . The ETD group had type C tympanogram with abnormally negative tympanometric peak pressure (TPP), which ranged from -155 to - 318 daPa. Their Yec and Ytm were within normal range, 0.6 - 0.9 mmho and 0.6 - 0.8 mmho, respectively. . The air conduction thresholds ranged from 5 to 75 dB HL (at 8 kHz) for the ETD group. The ER pattern in two adult cases (3 ears) with ETD is shown in figure 9. The ANOVA findings demonstrate statistically significant difference between the two groups (control adults and ETD group) only at 0.25 kHz (F [1, 18]) = 8.956, p = 0.00). This difference is mainly due to higher ER for the control group (0.993) than the ETD group (0.823) at that particular frequency (Table 6).

Upper left and right panel is the ER graph of a female adult's right and left ear with ETD. Lower panel shows the ER graph of a male adult's right ear with ETD.

Children subjects

The data for the WBER were analyzed using one way ANOVA to compare between the control children and the pathological groups (OME, TM perforation and ETD) as a function of octave frequency 0.25, 0.5, 1, 2, 4, and 8 kHz. A summary of the ER mean, SD, and ANOVA analysis between the children control group and the pathological groups: OME, TM perforation and ETD are shown in Tables 7, 8, and 9 respectively. Illustrations of the 10th, 25th, 50th, 75th, and 90th percentile of the energy reflectance of normal children and children with OME, Perforated TM and ETD are represented in figures 10, 11, and 12 respectively.

Control children vs. patients with OME. All control children had normal type A tympanogram with normal Yec (0.4 to 0.9 mmho) and Ytm (0.3 to 1 mmho), while children with OME had type B flat tympanogram with normal Yec (0.4 to 0.6 mmho). Pure tone Audiometry showed slight to mild hearing loss (mean = 25 - 40 dB HL) for the OME group compared to normal hearing in the control group (< 25 dB HL). Figure 10 demonstrates the ER pattern for children with OME, which is characteristic for such pathology. The overall ER was high across all frequency from 0.25 kHz to 4 kHz with a characteristic notch at 4.8 kHz and 6 kHz.

The ANOVA findings revealed statistically significant differences between the two children groups (control and OME) at 0.25 kHz (F [1, 26]) = 5.104, p = 0.03), 0.5 kHz (F [1, 26]) = 8.023, p = 0.009), 1 kHz (F [1, 26]) = 26.064. p = 0.000), and 2 kHz (F [1, 26]) = 6.817, p = 0.015). Children with OME had higher ER than the control group at 0.25 kHz (1.009 vs. 0.964), 0.5 kHz (1.006 vs. 0.908), 1 kHz (0.820 vs. 0.557), and 2kHz (0.579 vs. 0.355) (Table 7).

The lines represent the 10 th, 25 th, 50 th, 75 th, and 90 th percentile of the energy reflectance of the control children (left panel) and children with OME (right panel).

Control children vs. patients with TM perforation. Children with TM perforation had type B flat tympanogram with within normal Yec volume (2.5 to 7.2 mmho). Air conduction thresholds were < 25 dB HL for the control group and 25 dB HL, on average, for the TM perforation group. Compared to the control ER ratios, the mean ER in children with TM perforation was abnormally low across all frequencies, mainly 0.5 kHz and 4 kHz. Table 8 demonstrates the ANOVA results between the two children groups (control and perforated TM). The ANOVA findings reveal statistically significant differences between the two groups at 0.25 kHz (F [1, 27]) = 56.803, p = 0.000), 0.5 kHz (f [1, 27]) = 205.079, p = 0.000), and 2 kHz (F [1, 27]) = 4.635, p = 0.040). Children with TM perforation had lower ER at 0.25 kHz and 0.5 kHz (0.589 and 0.302, respectively) than the control group (0.964 and 0.908). At 2 kHz the reverse occurred; the ER for the TM perforated group was higher (0.521) than the control group (0.355).

The lines represent the 10 th, 25 th, 50 th, 75 th, and 90 th percentile of the energy reflectance of the control children (left panel) and children with perforated TM (right panel).

Control children vs. patients with ETD. Children with the ETD had type C tympanogram with normal Yec (0.6 to 0.7 mmho) and Ytm (0.7 to 0.8 mmho). Pure tone thresholds were < 25 dB HL for the control group, and 25 dB HL for the ETD group. Figure 12 shows a representation of the ER curve of a female child's right ear and a male child's left ear, both with ETD. A summary of descriptive statistics and the ANOVA results between the two children groups (control, n=22 and ETD, n= 2) is represented in Table 9. There was a statistically significant difference between the two groups only at 4 kHz (F [1, 21]) = 4.960, p = 0.04) due to higher ER for the ETD group (0.895) than the control group (0.425) at 4 kHz.

Left panel is the ER graph of a female childe right ear with ETD. Right panel shows the ER graph of a male childe left ear with ETD.

DISCUSSION

Wideband energy reflectance measurement has the potential to provide valuable information about middle ear function across wide frequency range and can distinguish abnormal from normal middle ear. This research is expanding the studies of WBER to investigate the middle ear function in normal and pathological conditions of the middle ear in adults and children. Overall, the study included 39 controls (17 adult ears and 21 children ears). The experimental groups with middle ear disorders consisted of 27 adult ears (10 Otosclerosis, 7 TM perforations, 7 Hypermobile TM, and 3 ETD) and 17 children ears (2 ETD, 8 TM perforations, and 7 OME). The relatively large numbers of cases make our study a more comprehensive than previously published work by Feeney et al. (2003) and Allan et al. (2005), who used less number of ears (13 in Feeney et al, and 4 in Allan et al) with middle ear pathological conditions.

Control groups (adults and children).

Normative data was collected from two control groups: adult group (13 male ears and 4 female ears, mean age = 35.2 years ± 10.8 SD) and children group (9 male ears and 12 female ears, mean age = 3.8 years ±1 SD). In the presence of normal pure tone thresholds, tympanometry results revealed that all adults and children control group had normal type A tympanogram with normal ear canal volume. The current finding of acoustic admittance values were within normal for the control adults (0.3 mmho to 1.7 mmho) (Margolis and Goycoolea, 1993), and the control children group (0.25 mmho to 1.05 mmho) (Roster et al, 2000).

The mean adults Energy Reflectance (ER) measured over one-six octave showed normal ER pattern. This pattern consisting of high ER in the low frequencies, decreased to a minimum [the most absorbed energy] between 1.5 kHz to 3.5 kHz, and then increased gradually at higher frequencies. Our finding as shown in Figure 5 also shows that most of the incident power is reflected back to the ear canal at frequencies below 1 kHz, whereas at mid frequency (1 kHz-5 kHz) of normal middle ear function most of the incident sound energy is absorbed by the middle ear. The reflected power below 1 kHz can be explained by the increased impedance mismatching at the TM (reactance is greater than resistance), while between 1 kHz-5 kHz the impedance of the TM is equal to the impedance of the ear canal (reactance is canceled), resulting in transmitting most of the sound energy to the middle ear (Allen et al, 2005). Using the same ER measuring equipment, several researchers have reported similar normal ER pattern in adults (Feeney et al, 2003; Keef et al, 1993; Margolis et al, 1999). In another study that used different ER equipment (Reflectance Measurement System IV [Mimosa Acoustics, Inc.]), Allen et al. (2005) tested a 50-year-old female with normal hearing. They also reported similar ER pattern (high ER at low frequencies, minimum ER level [40- 50%] between 1-3 kHz, and then high ER above 50% at higher frequency).

When comparing the ER pattern between the two control groups, our findings showed similar ER pattern between the two groups (Figure 5). However, the minimum ER extended over a wider frequency range (2 kHz - 3.8 kHz) in normal children compared to normal adults (1.5 kHz - 3.5 kHz). In addition, the children control group had a statistically significantly higher ER than the adult control group at 0.5 kHz and 1 kHz (Figure 5 and Table 2). Below 1 kHz, children middle ears have lower compliance than the adult ears (Okabe et al, 1988). This might explain the lower power transfer by the children middle ear at lower frequencies. Hunter et al. (2003) studied ER pattern in children ranging in age from 6 months to 4 years and found that the lowest ER was between 1.5 kHz to 4 kHz. Despite the age difference of children between the two studies, younger children in Hunter et al study compared to the current study, the ER results are still comparable. This normal pattern of ER supports the presence of normal type A tympanogram and, hence normal middle ear function.

Patient groups with middle ear pathologies

Adult patients with otosclerosis. ER was measured in female adults who have been clinically diagnosed with otosclerosis (mean age = 33.7 years ± 10.1 SD) (Table 1). The average pure-tone hearing thresholds were 51 dB to 60 dB HL and the average air-bone gaps for the otosclerotic group were 28.5 dB to 46 dB, which are consistent with mild to severe conductive or mixed hearing loss. Despite the clinical diagnosis of otosclerosis, all otosclerotic ears had normal 226-Hz type A tympanogram. Several investigators reported similar n