Signal to noise ratio as a guideline to obtain valid auditory brainstem response results

Dzulkarnain, Ahmad Aidil Arafat (2020) Signal to noise ratio as a guideline to obtain valid auditory brainstem response results. Malaysian National Society of Audiologist (MANSA) e-Bulletin 2020. pp. 20-21.

Preview

PDF - Published Version Download (1MB) | Preview

Abstract

Auditory brainstem response (ABR) is regards as one of the important tests to assess the auditory brainstem function and to estimate hearing thresholds. Whilst ABR is considered objective from the patient point of view, the interpretation of the ABR requires substantial knowledge from the perspective of an audiologist. The whole processes of ABR interpretations are complex. The first important aspect in interpreting the ABR finding is to identify whether an ABR is present or absent. Typically, the technique uses to identify a true ABR is by using response replication analysis. In this technique, ABRs are repeated twice under identical test conditions. Given a large area of overlapping between the two repeat ABRs, the clinician will have a high level of confidence that a true ABR is identified. This technique can also be used in conjunction with other techniques such as tracking the ABRs by observing the change in latency/amplitude as a function of intensity levels. In certain evoked potential system, response replication analysis is being provided statistically also known as cross-correlation. In cross-correlation, the recorded ABRs are separated into two recording buffers and the response from each buffer will be statistically compared. If the percentage of relationship are higher (e.g. more than 70%), it indicates a high probability that both recorded ABRs are highly correlated, thus increase the confidence level of the ABR detection. Although response replication is still considered as a valid method to identify ABR it poses at least one limitation. There is a potential that noise or artifact can mimic the morphology of the ABR and repeated at almost the same time or latency. This may lead to false identification and misdiagnosis of ABRs among the clinician. One of the recommended techniques to strengthen response replication analysis is to measure the recorded ABR signal-to-noise ratio (SNR). The ABR SNR can be computed manually or using certain type of analysis provided by the auditory evoked potential system. Example of automated SNR analysis include the F-ratio at single or multiple points (1). The F-ratio values will gradually increase when the estimated noise level from the selected points of the ABRs are lower than the amplitude of the ABR. Typically F-ratio of 3.1 is an indicator that the auditory evoked potential system is 99% confidence that the ABR is presence and well above the recorded noise level. Without this algorithm, clinician need to manually calculate the SNR. According to British Society of Audiology (2) a minimum acceptable limit of ABR amplitude is equivalent to 0.04 µV with a minimum ratio between ABR and the noise level of 3:1. The measurement of ABR amplitude is very straightforward by just simply plotting the peak of wave V and its following trough (at peak of SN 10). To estimate the noise, clinician need to observe the gap between two superimposed ABR waveforms and manually compute the noise amplitude based on the appropriate display scale (3). Therefore, clinician should only stop the ABR recording if the SNR is equivalent to 3:1 or higher (if the ABR is present). Some latest evoked potential system may provide on-going residual noise level to the clinician which can be used as a guideline. Figure 1 illustrates on the technique to identify the gap between two ABR waveforms and to estimate its residual noise level. Beside identifying ABR as present, clinician also need to decide when to accept ABR as absent. Based on own experience, question to decide whether ABR is considered as present or absent often raise because of the poor quality of a recording. Clinician may confuse whether to accept the ABR that contains substantial amount of noise versus a true ABR. Whilst this issue is not the aim of the paper, clinician should employed correct strategies to reduce the amount of background noise during the recording to ensure valid interpretations. An absence ABR is defined as a waveform that clearly flat with reasonably lower residual noise level that again can be assessed by observing the gap between two superimposed ABR waveforms or by the estimated residual noise levels provided by the equipment. The recommended minimum residual noise level to stop an ABR recording ranging from 0.025 µV to 0.04 µV (2).

Item Type:	Article (Bulletin)
Additional Information:	4465/86706
Uncontrolled Keywords:	Auditory Brainstem Response
Subjects:	R Medicine > RF Otorhinolaryngology
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Allied Health Sciences Kulliyyah of Allied Health Sciences > Department of Audiology and Speech-Language Pathology
Depositing User:	Prof Dr Ahmad Aidil Arafat Dzulkarnain
Date Deposited:	28 Dec 2020 12:16
Last Modified:	12 Apr 2022 08:43
URI:	http://irep.iium.edu.my/id/eprint/86706

Actions (login required)

View Item