Time-Frequency Properties of Neonatal Transiently Evoked Otoacoustic Emissions Measured in Three Different Acquisition Modes

Abstract

OBJECTIVE:The purpose of this study was to investigate transiently evoked otoacoustic emissions recorded from the cochleas of neonates using three acquisition protocols: non-linear; linear; and QuickScreen. In the non-linear mode, a series of four clicks was used, with three clicks at the same level and polarity and a fourth at three times greater amplitude and inverted polarity. In the linear protocol, all stimuli were presented at the same level and polarity. Both modes used a 20 ms recording window, but the third QuickScreen screening mode used the non-linear click sequence and a shorter 12 ms window. MATERIALS and METHODS:All responses were processed off-line with time-frequency algorithms, with the linear transiently evoked otoacoustic emissions additionally processed to eliminate linear artifacts in the initial few milliseconds of signal. RESULTS:Short windows contained only a fraction of the transiently evoked otoacoustic emission, distorting its time-frequency structure. Analysis of the response level and signal to noise ratio suggested that recordings elicited by a linear protocol had an advantage in terms of signal to noise ratio. CONCLUSION:The results indicate that screening protocols that use recording windows with a reduced length, such as QuickScreen, can only be used to detect the presence or absence of otoacoustic emission. Removing the stimulus artifact from linear transiently evoked otoacoustic emission responses improves response level and signal to noise ratio compared with the standard non-linear protocol.

Differences between Pressurized and Non-Pressurized Transient-Evoked Otoacoustic Emissions in Neonatal Subjects

“…Currently there is no consensus on what might constitute the required "extra" information. There are numerous schools of thought on this topic: the older school of thought postulates that important signal descriptors can be derived from different modalities of TEOAE signal decomposition and analysis (than the traditional fast Fourier transform) such as time-frequency analysis [Hatzopoulos et al, 2000[Hatzopoulos et al, , 2009a, time-scale analysis [Tognola et al, 1998;Jedrzejczak et al, 2012Jedrzejczak et al, , 2014, recurrence quantification analysis [Zimatore et al 2002[Zimatore et al , 2003Zimatore and Cavagnaro, 2015] etc. Despite the encouraging data presented in these studies, the results have not been translated into a clinical device or software.…”