Denoising Click-evoked Otoacoustic Emission Signals by Optimal Shrinkage

TL;DR

This paper introduces optimal shrinkage methods to improve the signal-to-noise ratio of click-evoked otoacoustic emissions, enabling more accurate and efficient cochlear function assessment with fewer measurements.

Contribution

It proposes covariance-based and SVD-based optimal shrinkage techniques to enhance CEOAE signal quality, outperforming traditional median-based methods in simulation and real data.

Findings

cOS improves SNR by 1-2 dB over baseline

sOS achieves 2-3 dB SNR enhancement in simulations

OS allows analysis of CEOAE dynamics over time

Abstract

Click-evoked otoacoustic emissions (CEOAEs) are clinically used as an objective way to infer whether cochlear functions are normal. However, because the sound pressure level of CEOAEs is typically much lower than the background noise, it usually takes hundreds, if not thousands of repetitions to estimate the signal with sufficient accuracy. In this paper, we propose to improve the signal-to-noise ratio (SNR) of CEOAE signals within limited measurement time by optimal shrinkage (OS) in two different settings: the covariance-based OS (cOS) and the singular value decomposition (SVD)-based OS (sOS). By simulation and analyzing human CEOAE data, the cOS consistently reduced the noise and enhanced the SNR by 1 to 2 dB from a baseline method (BM) that is based on calculating the median. The sOS achieved an SNR enhancement of 2 to 3 dB in simulation, and demonstrated capability to enhance the SNR in real recordings when the SNR achieved by the BM was below 0 dB. An appealing property of OS is that it produces an estimate of every individual column of the signal matrix. This property makes it possible to investigate CEOAE dynamics across a longer period of time when the cochlear conditions are not strictly stationary.