1. Distributed Sources of Otoacoustic Emissions
Stephen Neely
Michael Gorga
Boys Town National Research Hospital
Omaha, Nebraska

2. Story of OAE sources History of OAE-source theories
Observations of multiple temporal peaks
DPOAE IFFT
Cochlear reflectance
Clinical utility of DPOAEs
Benefits of multivariate analysis
DPOAE suppression

3. Where do OAEs originate?
Kemp (1978)
Kim, Siegel, Molnar (1979)
Neely, Norton, Gorga, Jesteadt (1988)
Prieve, Abbas (1989)
Zweig, Shera (1995)
Stover, Neely, Gorga (1996)
Talmadge, Tubis, Long (1998)
Shera, Guinan (1999)
Konrad-Martin et al. (1999)
Choi et al. (2003)

4. 1978: Kemp
Stimulated acoustic emissions from within the human auditory system
First OAE measurements
“…probably of cochlear origin”

5. 1979: Kim, Siegel , Molnar
Cochlear nonlinear phenomena in two-tone responses
2f1-f2 distortion component is present
at the f2 place
at the 2f1-f2 place
Neural data (not a cartoon)
As a graduate student in Kim’s lab, these data strongly influenced by thinking about otoacoustic emissions.

6. 1988: Neely, Norton, Gorga, Jesteadt
Latency of auditory brain‐stem responses and otoacoustic emissions using tone‐burst stimuli
ABR latency remarkably consistent
OAE latency twice ABR, but difficult to define
Cochlear origin!

7. 1995: Zweig, Shera
Modeling otoacoustic emission and hearing threshold fine structures
Introduced coherent reflection theory.

8. 1996: Stover, Neely, Gorga
Latency and multiple sources of distortion product otoacoustic emissions
Pioneered IFFT analysis of DPOAEs
Associated IFFT peaks with sources
Sometimes more than two peaks

9. 1998: Talmadge, Tubis, Long
Modeling otoacoustic emission and hearing threshold fine structures
Developed theory of two-source model

10. 1999: Shera, Guinan
Evoked otoacoustic emissions arise by two fundamentally different mechanisms: A taxonomy for mammalian OAEs
Two mechanisms
Linear reflection
Nonlinear distortion
Two broad classes of OAEs
Reflection source
Distortion source
Taxonomy widely accepted and influential

11. 2001: Konrad-Martin, Neely, Keefe, Dorn, Gorga
Sources of distortion product otoacoustic emissions revealed by suppression experiments and inverse fast Fourier transforms in normal ears
Further developed IFFT analysis of DPOAEs
Observed multiple DPOAE peaks
Demonstrated suppression of secondary DPOAE peaks
Associated unsuppressed peaks with distortion sources and suppressed peaks with reflection sources

12. 2008: Choi, Lee, Parham, Neely, Kim
Stimulus-frequency otoacoustic emission: Measurements in humans and simulations with an active cochlear model
Salient features of human SFOAEs were simulated with an active cochlear model.
“…in the model, an SFOAE consists of a long-delay component generated by irregularity in the traveling-wave peak region and a short-delay component generated by irregularity in cochlear regions remote from the peak.”

13. Summary of OAE sources Consensus on cochlear origin
SFOAEs contributions probably from multiple locations
DPOAE contributions definitely from multiple locations: stimulus place, distortion place, and (possibly) basal locations
Multiple sources represent not a problem, but an opportunity

14. Multiple Temporal Peaks
DPOAE inverse FFT (Stover et al., 1996)
Cochlear contribution to time-domain ear-canal reflectance (Rasetshwane & Neely, 2012)

15. DPOAE IFFT f1 sweep with fixed f2 Inverse FFT reveals multiple peaks
(Stover et al., 1996)
f1 sweep with fixed f2
Inverse FFT reveals multiple peaks

16. DPOAE IFFT: Peak latency vs. level
Latency of individual peaks remains constant
Peaks with longer latency become relatively smaller
Group delay decreases
(Stover et al., 1996)

17. Cochlear reflectance Cochlear contribution to ear-canal reflectance
Wide-band noise stimulus
IFFT -> waveform
Similar to SFOAE, but normalized to forward pressure
(Rasetshwane & Neely, 2012)

18. Cochlear reflectance Bandpass filtered (4 kHz)
Latency of temporal envelope peaks remains constant as stimulus level increases
(Rasetshwane & Neely, 2012)

19. Cochlear reflectance: Peak latency vs. level
Latency of individual peaks remains constant
Peaks with longer latency become relatively smaller
Group delay decreases
(Rasetshwane & Neely, 2012)

20. Summary of temporal peaks
DPOAEs and SFOAEs similarly possess multiple peaks in temporal envelopes with intensity-invariant latency
Group delay decreases because relative amplitudes of peaks with longer latency decrease
Temporal peaks may be associated with distinct locations within the cochlea

21. Clinical Utility of DPOAEs
Shaffer et al. (2003)
Dorn, Piskorski, Gorga, Neely, Keefe (1999)
Kirby, Kopun, Tan, Neely, Gorga (2011)
Thorson, Kopun, Neely, Gorga (2012)
Gorga et al. (2011)
Johnson et al. (2007)
Neely, Gorga, Kopun, Tan (2011)

22. 2003: Shaffer, Withnell, Dhar, Lilly, Goodman, Harmon
Sources and Mechanisms of DPOAE Generation: Implications for the Prediction of Auditory Sensitivity
“The interactions of multiple sources of OAEs arising by distinct mechanisms complicate the simple interpretation that OAE frequencies and hearing test frequencies assess the integrity of the same cochlear locations.”

23. 1999: Dorn, Piskorski, Gorga, Neely, Keefe
Slow down. The parameter is F2. Example F2=2 kHz.
DPOAE amplitude depends on hearing status at frequencies above f2
Suggests that DPOAEs have basal contributions

24. Predicting Audiometric Status from DPOAEs Using Multivariate Analyses
Information from DPOAE measurements at multiple f2 frequencies can be combined via multivariate analysis
Pass/fail test performance is improved by multivariate analysis
(Dorn et al., 1999)

25. Do “Optimal” Conditions Improve Distortion Product Otoacoustic Emission Test Performance?
Eleven years later, Kirby et al. (2011) confirmed the benefit of multivariate analysis on pass/fail test performance
Still no commercial implementation of multivariate pass/fail decisions

26. 2012: Thorson, Kopun, Neely, Gorga
Multivariate analysis of DPOAE I/O functions have been shown to predict categorical loudness functions

27. DPOAE suppression tuning curves
(Gorga et al., 2011)
Group means eliminate variability due to multiple sources
DPOAE STC quality is outstanding

28. Attempts to eliminate “reflection source” from DPOAE measurements have been disappointing
Johnson et al. (2007): “…controlling source contribution with a suppressor did not improve diagnostic accuracy and frequently resulted in poorer test performance compared to control conditions.”
Neely et al. (2011): “…averaging DPOAE measurements across closely spaced f2 frequencies not an effective way to reduce within-subject variability.”

29. Summary of Clinical Utility
Multivariate analysis of DPOAEs at multiple f2 frequencies improves pass/fail test performance
Multivariate analysis of DPOAE I/O functions may predict categorical loudness
DPOAE suppression tuning curves best available
Attempts to eliminate reflection source have been disappointing

30. Conclusions
Multiple locations in the cochlea probably contribute to both DPOAEs and SFOAEs
Distinction between distortion source and reflection source is useful in theory, but less useful in practice
Elimination of reflection source from DPOAEs not necessary to achieve high-quality results:
Pass/fail test performance
Tuning curves
Loudness prediction

31. Acknowledgements Co-authors Funding from NIH/NIDCD
Grants R , R , and P

32. …

33. 2007: Johnson et al.
Distortion product otoacoustic emissions: Cochlear-source contributions and clinical test performance
“…controlling source contribution with a suppressor did not improve diagnostic accuracy and frequently resulted in poorer test performance compared to control conditions.”

34. 2011: Keefe et al.
Detecting high-frequency hearing loss with click-evoked otoacoustic emissions
Various OAE types have similar pass/fail test performance