Elsevier

Hearing Research

Volume 231, Issues 1–2, September 2007, Pages 54-62
Hearing Research

Research paper
Otoacoustic emissions latency difference between full-term and preterm neonates

https://doi.org/10.1016/j.heares.2007.05.009Get rights and content

Abstract

Transiently evoked otoacoustic emissions (TEOAEs) were recorded from full-term and preterm neonates. The responses were decomposed, by means of an adaptive approximation method, into waveforms of defined frequencies, amplitudes, latencies and time spans. Statistically significant differences in the latency values were found between the tested groups. Differences were also found in the time spans of the TEOAEs components. For the preterm neonates the contribution of long-duration components (i.e. long-time span) was higher. Those components were characterized by narrow frequency band and contrary to the short-time span components their latencies did not depend on frequency. The removal of the long-duration components, from the pool of analyzed data, decreased the latency differences between the tested groups. The results indicate that the origin of the longer latency values for preterm neonates (with a post conceptional age up to 33 weeks) in respect to full-term neonates can be attributed to the presence of long-lasting components. The correspondence, which was found between frequencies of long-duration components and the spectral peaks of spontaneous otoacoustic emissions (SOAEs), suggests that those components may be connected with SOAEs.

Introduction

Otoacoustic emissions (OAEs) are low level sounds of cochlear origin and can provide information on the functional status of the cochlea. OAEs have been particularly useful in detecting hearing deficits in neonates, children and adults as a non-invasive test, not requiring the cooperation of the patient (Brass and Kemp, 1994, Aidan et al., 1997, Smurzynski et al., 1993, Stevens et al., 1990). They arise as a spontaneous and evoked activity and in the second category according to a stimulus paradigm the transiently evoked otoacoustic emissions (TEOAEs) and the distortion product otoacoustic emissions (DPOAEs) are commonly placed (Probst et al., 1991).

Usually the discrimination between TEOAE responses from normal-hearing subjects and patients with sensorineural hearing losses is based on the reproducibility (or correlation) of the TEOAE response, the signal to noise ratio (S/N) and overall TEOAE response level (e.g. Hatzopoulos et al., 1995, Moleti et al., 2002). More recently other TEOAE parameters have proven to be good signal descriptors and among those the latencies of the TEOAE components. It was demonstrated that there are differences in latencies of OAEs between subjects with normal hearing and subjects exposed to noise (Sisto and Moleti, 2002, Jedrzejczak et al., 2005). The TEOAE latency-frequency relation may also provide an estimate of the cochlear tuning.

Otoacoustic emissions can be also a tool to study development of active processes in cochlea and its maturation. These studies concerned various classes of OAEs such as: spontaneous emissions (e.g. Brienesse et al., 1998, Morlet et al., 1995) or evoked otoacoustic emissions (e.g. Chuang et al., 1993, Smurzynski, 1994, Brienesse et al., 1996). Recent studies on the subject of cochlear maturation and OAEs have used wavelet analysis on the TEOAE responses (Tognola et al., 2005, Moleti et al., 2005). In their study of preterm and full-term neonates Tognola et al. (2005) have shown that the TEOAEs latencies of the preterm subjects are longer than those of full-term subjects. To explain these observations several hypotheses have been proposed, which have considered as major contributors to the observed latency differences, the immaturity of the cochlea and/or the immaturity of the middle ear in the preterm neonates. Lower absorption of the energy in the immature middle ear in comparison with adult ear, as shown by Keefe et al. (1993), was connected with possible increase of latency.

In this paper we continue previous studies of latency of OAEs in neonates by using more advanced signal processing methodology namely adaptive approximations by the matching pursuit (MP) approach. The MP procedure decomposes in an adaptive way the TEOAE signal into components (or modes) chosen from very large dictionary of functions. In this context the signal is represented by the waveforms of the well-defined frequencies, latencies and time spans. The description of the TEOAE signal in terms of well-defined parameters (more efficient descriptors) creates the possibility of looking deeper into the properties and structure of the TEOAE responses. By means of MP approach (Jedrzejczak et al., 2004, Jedrzejczak et al., 2005), it was demonstrated that the TEOAE signal can be considered as a superposition of the resonant modes of characteristic frequencies and latencies.

The present paper considers not only the TEOAE latencies but also the time spans of the TEOAE components. The time span is a new parameter in OAE analysis and it is related to the duration of OAE component of defined frequency.

Section snippets

Datasets

TEOAEs from preterm and full-term neonates were measured in low noise ambient conditions by a ILO-292 apparatus (Otodynamics LTD, London) running software version 5.6. The full-term neonates were also tested otoscopically for any external ear abnormalities (i.e. excessive wax, collapse of the meatal canal, etc). Standard non-linear TEOAE protocols were used for the collection of data (average stimulus of 80 peak dB SPL; 260 repetitions). The TEOAE signal bandwidth was set from 0.8 to 6.0 kHz

Results

The representation of the TEOAE signal, in terms of components of known energy, latency and time span, provides parameters suitable for statistical analysis and allows the construction of the time–frequency (TF) distribution of the TEOAE signal energy. Examples of TEOAEs TF maps for preterm and full-term neonates are presented in Fig. 2. At the bottom of each panel the first seven resonance modes of the TEOAE signal (in descending order of energy) are shown. Inspecting the TF distributions it

Discussion

Here the factors contributing to the latency difference between full-term and preterm neonates were investigated. Latencies and time durations of the single frequency components were determined. The high time–frequency resolution of the matching pursuit method, allowed the identification of the two kinds of waveforms (components) appearing in the TEOAE responses. The contributions of these components were different across the three studied groups. This effect turned out to be important for

Acknowledgement

This work was supported by statutory grant of Ministry of Higher Education and Science to Institute of Experimental Physics of Warsaw University.

References (32)

  • D. Brass et al.

    The objective assessment of transient evoked otoacoustic emissions in neonates

    Ear Hear.

    (1994)
  • P. Brienesse et al.

    Otoacoustic emissions in preterm infants: indications of cochlear development?

    Audiology

    (1996)
  • P. Brienesse et al.

    Long-term and short-term variations in amplitude and frequency of spontaneous otoacoustic emissions in preterm infants

    Audiology

    (1998)
  • E.M. Burns et al.

    Longitudinal measurements of spontaneous otoacoustic emissions in infants

    J. Acoust. Soc. Am.

    (1994)
  • J. Cheng

    Time–frequency analysis of transient evoked otoacoustic emissions via smoothed pseudo Wigner distribution

    Scand. Audiol.

    (1995)
  • S. Hatzopoulos et al.

    Identification of hearing loss using TEOAE-based descriptors: preliminary results

    Audiology

    (1995)
  • Cited by (0)

    View full text