Future Directions of Basic and clinical OAE research

From the Board of the OAE Portal (edited by Stavros Hatzopoulos)

Department of Audiology , Uiversity of Ferrara, Italy

Introduction

          In 2005 Otoacoustic Emissions celebrate a life span of 28 years (after the first OAE publication by David Kemp), but despite this long time period numerous areas of basic and clinical OAE research have remained partially unexplored and undefined.

           This report  ( a compilation of information presented already in the Portal plus new material) will provide information on the latest trends in the OAE field and will present information about areas which need to be further explored.

Biophysics, Cochlear Mechanics and generation mechanisms of OAEs.

The issue of cochlear mechanics related to the OAE generation is always a quite complex domain for the majority of clinical and health professionals. Nevertheless the understanding of the genesis of OAEs is the main channel that can lead to new clinical applications. The last 5 years considerable effort has been dedicated to new cochlear models which can explain accurately the OAE generation from low and high stimuli (Nobili et al , 1998; 2003; Shera 2003; 2004; Shera et al, 2000; 2002; 2004 ) . Following the traditional clinical categorization  (ie OAEs categorized by the invoking stimulus) we had to construct mathematical models and perform computer simulations to explain how an acoustic stimulus can give raise to transient otoacoustic emissions (TEOAEs) and distortion product otoacoustic emissions (DPOAEs). Once we had these two classes explained then we could tune the model to explain the generation of spontaneous otoacoustic emissions (SOAEs).

            In 1999 Shera and Guinnan offered a revised version of the OAE classification. The argued that the OAEs are the sum of linear reflections (giving raise to TEOAEs) and nonlinear frequency distortions (giving raise to DPOAEs). Although this classification is closer to the physiology of the auditory system it creates considerable confusion among Audiology students and ENT residents, because in the proposed model each OAE response contains both TEOAEs and DPOAEs. The stimulus level is the determining factor of which type prevails over the other.

            At present there are two main theories which tend to explain the genesis of TEOAEs. The first theory by Christopher Shera (Shera 2003; 2004; Shera et al, 2000; 2002; 2004 ) assumes that at the peak of the travelling wave a part of the wave's energy is reflected for reasons of geometry and wave-dynamics (linear reflection). At the same time the high level dynamics of the cochleal amplifier introduce intermodulation distortion (ie the genesis of the DP families of responses). Another model proposed by Nobili et al (2003 )  assumes that the OAEs  are generated by imperfections in the middle ear tranfer function. Comments on both models can be found in the Portal sections.

             DPOAEs have been studied in more detail because it was found that the recoded cubic distortion product (ie 2f1-f2) is very sensitive to the presence of higher distortion products which act as suppression tones.  In the mid-nineties while  the clinical research of DPOAEs was peaking, some predictions were made regarding the extrapolation of useful cliical information from the plethora of distortion product families (ie 3f1-f2, 4f1-2f2, etc). Recent reseach ( Fahey et al, 2000; Gorga et al, 2002;Martin et al, 2003 ) has indicated that higher order distortion products are subjects to significant suppression effects (similar to the well-known  2-tone suppression interaction) and for this reason the amplitudes of these products are very low  and therefore incable of helping us with a more precise clinical assessment of the cochlear status. Nevertheless the relationship between these families of DPOAEs and various contralateral suppression effects  is still un-exlored and it might be that these products can help us understand better central interactions with the auditory periphery.

               One of the future goals in biophysical research and OAEs is to define better and faster models which can provide more accurate descriptions of the cochlear functionality, which is suppressed and  damaged after  administration of ototoxic substances or after exposure to high sound or noise levels. Currently the protocols employed are requiring too much linear information (ie long recording times, simple TEOAE or DPOAE protocols). Combinational protocols  (various types of TEOAEs + DPOAEs) offered by fast acquistion devices might provide the necessary information.

New (and possibly clinical) indices which can be extrapolated from the current protocols (ie TEOAE and DPOAE) to serve better inner ear assessment.

           Traditionally the OAE responses are interpreted in the frequency domain and the majority of the PASS criteria are based on the estimation of signal to noise ratios (S/N)  at 2.0 , 3.0 and 4.0 kHz. The presence of a good (ie statistically acceptable) S/N ratio at a specific band is considered a good indicator that at that band the hearing theshold is almost at normal levels. If the data belong to an infant an acceptable S/N ratio is an indicator of a threshold better (ie lower) than 30 dB HL.

            Inner Ear assesment via the OAEs has been primarily focused in the area of neonatal hearing screening , although other clinical areas such as OAE ototoxicity monitoring are becoming more popular in the clinical practice. OAEs can tell us something about the cochlear functionality (ie normal or abnormal), but the information we get from traditional protocols stops there.

            Recently a number of authors (Jannsen et al, 1995;, Kummer et al, 2000; Dorn et al, 2001; Boege and Jannsen; 2002; Gorga et al, 2002; 2003; Neely et al, 2003; ) has proposed to use DPOAE measurements near the hearing threshold , in order to produce specific Input &Output  (IO) curves , from which  a hearing threshold estimation could be derived. One of the experimental challenges that the above authors have resolved is related to the fact that in order to get a threshold estimation measurements with very low stimuli must be made. But at near-to-threshold (ie at low primary tone levels) either no DPOAEs or DPOAEs with unsufficient signal-to-noise ratios can be measured. Therefore, when plotting the DPOAE level Ldp as a function of  f2 (in the DP-gram) the DPOAEs often do not reflect cochlear hearing thresholds. This issue was resolved by using information from extrapolated DPOAE I/O-functions to estimate hearing  threshold. In order to attain this sort of information we need to know the relationship between the DPOAE level Ldp and the primary tone level L2.

            Janssen et al. (1995a,b) and Kummer et al., (2000 have proposed a primary tone level setup, in which the difference between L1 and L2 increases with decreasing stimulus level. Using this paradigm, (known also as an asymmetrical DPOAE protocol ie 60-50, 70-60 dB SPL) instead of the commonly used equal level paradigm (ie 65-65 dB SPL) the DPOAE growth reflects the compressive nonlinear cochlear sound processing, which is known from direct measurements of basilar membrane motion in animal experiments (Ruggero et al., 1997; Boege and Janssen, 2002 ). A graph (see Figure 1 ) relating the amplitudes of L1, L2 and the amplitude of the cubic distortion product produces a scissor-like shape and this setup was coined the scissor-paradigm. 

Figure 1: The scissor paradigm (from Boege and Jannsen, 2002 ). The shape defined by L1, L2 and the amplitude of the cubic distortion product shows a conical form which resembles a pair of scissors pointing upwards. According to this set-up the most robust  DPOAE responses are obtained by asymmetrical protocols ( L1 > L2) , but this advantage disappears abole intensities of 65 dB SPL . From that level asymmetrical and symmetrical protocols generate equal amplitude DPOAE responses.

          Using this paradigm with  L1=0.4L2+39dB, most of the DPOAE I/O-functions recorded in normal-hearing human ears,  demonstrate a logarithmic dependency of the distortion product sound pressure level LDP on the sound pressure level L2 of the f2 primary tone ( Boege and Janssen, 2002). In normal-hearing ears, in the low primary tone level range, the slope approximates the value 1.0 dB/dB, whereas in the high primary tone level range the slope reaches the value of  0.333  dB/dB. With increasing hearing impairment  the slope value continuously increases ( Janssen et al., 1998; Kummer et al., 1998). This pattern of DPOAE growth is quite similar similar to the behavior described by direct basilar membrane measurements ( Ruggero et al., 1997). Thus, the DPOAE I/O-functions are able to reflect the compressive sound amplification in the cochlea at the outer hair cell level.The logarithmic dependency of the DPOAE sound pressure level on the primary tone sound pressure level, results in a linear dependency between the DPOAE sound pressure pDP and the primary tone sound pressure level L2.

           According to the data presented by Boege and Janssen, (2002) a close correspondence has been  found between the estimated cochlear pure-tone threshold and the behavioral threshold recorded with the same sound probe.  When comparing the behavioral pure-tone threshold LT and the estimated cochlear pure-tone threshold LEDPT for 4236 DPOAE I/O-functions of 30 normal-hearing and 119 cochlear hearing loss ears that fulfill linear regression criteria (for detail see Boege and Janssen, 2002) a significant correlation was found present. Moreover, there is almost a 1:1 relationship between the subjective and the objective measures. This means that there is a direct quantitative relationship between the estimated cochlear pure-tone threshold and the behavioral pure-tone threshold.

           Recently, Gorga et al. (2003) extended this method by increasing the primary tone level (up to 85 dB SPL) and changing the criteria for accepting I/O-functions. The Gorga study replicated the results from (Boege and Janssen, 2002) when using the same stimulus conditions and the same linear regression criteria. Taking measurements for a wider range of levels and slightly altering the inclusion criteria Gorga et al. achieved an improvement in the test performance. They found prediction errors not to be uniformly distributed across the test frequencies. The best performance was observed in the mid-to-high frequencies. In a retrospective study (Oswald and Janssen, 2003) on the data reported by Boege and Janssen in 2002 using weighted extrapolated DPOAE I/O-functions similar results were obtained which were attributed to the in-the-ear-canal sound pressure calibration Therefore, further efforts are necessary to improve the sound pressure calibration in the outer ear canal for applying definite sound pressure at the ear drum and hence improving cochlear pure-tone threshold estimation. It should be noted that the precision of the threshold estimation is based on the data collected from adult subjects and extrapolated for general use (i.e. children and infants). One of the sensitive parameters is this procedure are the criteria for accepting the validity of the I/O functions. Gorga et al have also raised this issue in several papers (2002a, 2002b, 2003).

           Besides the estimation of pure-tone thresholds, DPOAE I/O-functions provide an additional measure. That is the slope of the I/O-function, which is able to estimate the compression of outer hair cell amplifier. This was shown in guinea pigs in which the outer hair cells were impaired using acute furosemide intoxication (Mills and Rubel, 1996) and for humans suffering from cochlear hearing loss (Janssen et al., 1998; Kummer et al., 1998; Boege and Janssen, 2002; Neely et al., 2003>). In these studies the slope of the DPOAE I/O-function increases with increasing hearing loss revealing loss of compression of outer hair cell amplifiers. In the future and with OAE devices equipped with fast processors, it might be possible to extrapolate very precise information related to the status of the cochlear amplifier by using numerous combinations of I/O  DPOAE recordings.

            An important problem in neonatal hearing screening is the interpretation of the middle-ear effect on the OAE measures. Recently, extrapolated DPOAE I/O-functions were constructed from human neonates to estimate cochlear pure-tone threshold and compression estimates (Janssen et al., 2003). The estimated pure-tone threshold was found to be increasing within the early postnatal period (average age: 3 days), predominantly at the higher frequencies, and to be normalized in a follow-up measurement (after four weeks). However, the slope of DPOAE I/O-functions obtained in the first and second measurement was unchanged revealing normal cochlear compression. Consequently, these findings were interpreted as temporary sound conductive hearing loss due to the presence of amniotic fluid and/or Eustachian tube dysfunction. In this clinical scenario, especially during the first days of life, a hearing screening test may lead to false positive results due to a temporary sound conductive hearing loss. The use of the slope of DPOAE I/O-functions could be used as an index of conductive losses which might result in less false positives an in less time spent for audiological diagnostic testing. To note that the DPOAE slope methodology can differentiate between a conductive impairment and a sensorineural.

           Another potential application of extrapolated DPOAE I/O-functions is to objectively adjust hearing aids in children. Since DPOAE I/O-functions are reported to be correlated with loudness (Neely et al. 2003), DPOAE would also offer the potentiality of basic hearing aid adjustment.

         The research findings from Janssen et al (2003) have been commercialized in a device called Cochlea- Scan by Fischer-Zoth. Figures 2 and 3 show the device and data acquisition sequences. At present the Cochlea-Scan device offers a platform for a third generation OAE testing (TEOAEs, DPOAEs), I/O DPOAE estimation with hearing threshold extrapolation and Pure Tone Audiometry measurements. Nevertheless measurements in the NICU environment should be always accompanied by additional AABR / ABR testing to ensure lack of retrocochlear pathologies, common in the NICU population (ie Auditory Neuropathy).

Figure 2 : The Cochlea-scan device by Fischer-Zoth. With the current firmware (ie May 2005 release) it is possible to do DPOAE screening, estimation of hearing thresholds up to a relative 50 dB HL and traditional Pure Tone Audiometry testing (external headset is required).

Figure 3: Cochlea-scan displays during the threshold estimation (top panel) and threshold final extrapolatiom (bottom panel) from a neonatal subject .The audiometric notch between 2 .0 to 4.0 kHz is enhanced by the standing waves in the external auditory meatus.

More complex / accurate methods of OAE analysis.

             In order to interpret the OAE data, the time-domain responses are transformed into the frequency domain and numerical parameters related to various frequency bands are calculated. These numerical estimates are used as normality criteria, PASS criteria, REFER criteria etc. Unfortunately the basic premises of the frequency transformation (i.e. when using the Fast Fourier Transform -FFT) are violated because the OAEs have characteristics which change every few ms (they are not stationary biosignals).

             In this context, the last 7 years numerous complex signal-processing schemes have been applied to the analysis of OAE signals. These schemes have the advantage that they do not possess the weaknesses of the FFT approach. Within the proposed methodologies the most important are: the Time-scale analysis, better known as wavelet analysis (Tognola et al , 2000; 2005 ); the Time-Frequency analysis (Hatzopoulos et al , 2000a, 2000b; Jedrzejczak et al, 2004); the Recurrence Quantification Analysis<Zimatore et al, 2002, 2003); and the Maximum Length Sequences / or the Voters series (Thornton ARD et al, 1997; 1999; 2001 ). All these methodologies have provided new insights mostly into the complicated nature of TEOAEs, but no results, which could be extended into new clinical applications.

           For example using Time-Scale analyses we have verified the maturation processes presented in the Inner ear as pre-term subjects age from 30 to 60 weeks (Tognola et al, 2005). Data from Time-Frequency analyses (Hatzopoulos et al, 2000) have indicated that the TEOAEs responses contain impulsive components, modulated frequency components (probably related to DPOAEs) and fixed-frequency components (probably related to SOAEs). Data from the Voltera series (Thornton et al. 2001) have suggested that we might have more sensitive tools in detecting the non-linearities of the cochlear amplifier and applications related to hearing screening and ototoxicity monitoring (mainly from noise) might benefit significantly.

Alternative Technologies:  the relationship between OAEs, Automated ABR (AABR) and Auditory Steady State Responses (ASSR).

          In the early 2002, the first 4rth generation OAE devices appeared in the market and provided the possibility to numerous clinical realities to integrate automated OAE (AOAE) and automated ABR (AABR) recordings. This scenario targeted the identification of auditory neuropathy cases most prevalent in the NICU environment. For the first time OAE recordings were directly compared with ABR recordings and from these comparisons several conclusions were reached (see the white paper section in the OtoAcoustic Emission Portal) , such as : (i) the average time for a  AOAE responses is clearly less than 10 s in a cooperative subject, and less that 120 s (2 min) in non-cooperative subjects. (ii) test times of AABR in cooperative subjects were less than 120 s , while uncooperative subjects were tested within 10 min (per ear).  The testing times between the two methodologies are getting very close, but the flexibility offered by AOAEs is difficult to be matched. While it takes some training to place the OAE probe correctly, the ABR electrode placement presents more complications especially in cases where the subject shows high electrode impedance. In the latter case the AABR testing difficult to complete.

         Both OAE and ABR technologies utilize as stimuli electrical clicks and the acquired information is clearly more related to the audiometric frequencies of 1.0 and 2.0 kHz. Within this context, there has been a speculation of whether other similar technologies could be used in the evaluation of the hearing threshold in neonates, children and adults. Other electrophysiological measurements involve EcoG, Middle latency and Steady State Responses (SSR). From this group the latter category has shown interesting characteristics due to fact that by alternating the modulation frequency (i.e. increasing it)  of the stimuli one can get responses from the Auditory cortex (low modulation frequencies around 40 Hz)  or from the Brainstem (Cone-Wesson et al; 2002; Dimitrijevic et al , 2002: John and Picton, 2002). The SSR protocol has already passed to an automated one (ASSR) and for the last two years numerous publications have been devoted to the threshold estimation via the ASSR technique. The ASSR protocols have been greatly optimized, (Gorga et al, 2004) and the SSR responses are detected in the frequency domain by robust probabilistic algorithms.

        Considering the above mentioned  factors it is not surprising that in 2002  Conne-Wesson and et al proposed the use of ASSR as a hearing screening tool, with the objective that ASSR could substitute the AABR. A few  reports have been available since (Stueve and OÕRourke, 2003; Luts et al, 2004;Swanepoel et al; 2004 ) indicating a good agreement between ASSR and AABR at 2.0 kHz and various differences at 0.5, 1.0 and 4.0 kHz. Most studies recommended the use of the technique in the clinic but the point of substituting the AABR with ASSR has not arrived yet. The factors which affect the AABR (ambient noise and electrode impedance) interfere with the ASSR recordings as well. Recently Vivosonic has presented a new line of devices using preamplifiers at the level of the scalp-electrodes (called amplitrodes) which suppress the level of ambient noise and provide very clean AABR and ASSR traces. It is to be seen how these electrodes will be intergraded in the normal clinical reality (the pre-amplifiers require electrical energy which translates into changing batteries every x tests !!).

Trends in the OAE Industry

            The innovation in the OAE field has shown a particular trend the last 20 years. It is not the clinician who defines his needs , but it is the OAE industry which defines what the clinician and researcher might do with OAEs. For example the contralateral suppression method , initiated in the early 90s by Lionel Collet, has not been established to the majority of OAE devices. To note that contralateral suppression is the only tool we have at our disposal  to probe accurately central hearing impairment , specifically residing in the cochlear nuclei.

         The last few years the industry trend was to pass from third generation to fourth generation devices (i.e. AOAE + AABR) .  Some manufacturers such as GN-Otometrics and Etymotic Research have mentioned projects integrating OAEs and impedance testing but no products have been announced in the 2005 Conference of the American Academy Of Audiology in Washington DC. Similarly reports from Fischer & Zoth indicate the trend to encapsulate on one small device many testing protocols not only for screening but for diagnostic purposes as well (TEOAEs, DPOAES, Cochlea Scan, PTA and probably ASSR). These trends underline the possibility that in the future it will be feasible to use protocols offering more information about the tested subjects which can be used either for screening purposes or for diagnostic statistics.

            The future of any electrophysiological method, which presently serves as a screener protocol, seems to be rather limited. The tremendous advantages in genetics research and the identification of numerous genes responsible for many types of hearing impairment (syndromic losses mainly or impairment due to specific diseases such as otosclerosis) are promising a near-by future where the majority of hearing testing will be conducted via genetic probes from a small blood sample. At present the American company Nanogen (www.nanogen.com) has released specific kits which can identify over 8000 gene mutations related to the connexin family (related to syndromic hearing impairment for the moment). It is expected that in the next 4 years the number of identified mutation will be more than 50000 (unpublished data from the latest conference on the Genetics of Deafness (GENDEAF), Caserta Italy, 2005).

References

Boege P, Janssen T. Pure-tone threshold estimation from extrapolated distortion product otoacoustic emission I/O-functions in normal and cochlear hearing loss ears.  J Acoust Soc Am 2002 111 (4) 1810-1818.

Cone-Wesson B, Dowell R, Tomlin D, Rance G, Ming Wu. The Auditory Staedy State Response. Comparisons with the Auditory Brainstem Response.  J Am Acad Audiol 2002; 13: 173-187.