"Linear" and "derived" otoacoustic emissions in newborns: a comparative study

Area of Science:

• Audiology and Otoacoustic Emissions
• Neonatal Hearing Screening
• Signal Processing in Bioacoustics

Background:

• Click-evoked otoacoustic emissions (CEOAEs) are a non-invasive measure of cochlear outer hair cell function.
• Acquisition techniques, specifically averaging methods, can influence the reliability and interpretation of CEOAE measurements.
• Understanding these influences is crucial for accurate neonatal hearing screening.

Purpose of the Study:

• To compare the effects of two averaging techniques (linear and derived) on CEOAE evaluation in newborns.
• To assess how averaging method impacts signal-to-noise ratio (SNR), reproducibility, and screening pass/fail classification.

Main Methods:

• CEOAEs were recorded from 89 newborns using an Otodynamic ILO88 system.
• Two averaging modes were applied: linear (classic) and derived (artifact cancellation).
• CEOAEs were analyzed for waveform similarity, signal/noise levels, inter-test reproducibility, and pass/fail status based on clinical criteria.

Main Results:

• Derived CEOAEs exhibited higher residual noise (6 dB) and lower inter-test reproducibility compared to linear averaging.
• Waveform differences and time shifts were noted at longer latencies between the two methods.
• Averaging technique significantly influenced pass/fail scoring, with 14% of linear 'pass' results becoming 'fail' with derived averaging.

Conclusions:

• The derived averaging technique results in noisier CEOAEs in newborns compared to the linear technique.
• Screening criteria relying on SNR can be affected by the chosen averaging method.
• Careful consideration of the averaging technique is necessary for consistent and accurate neonatal hearing screening.