Progress on the ANSI AEP Standard
Robert Burkard, PhD, CCC-A
The American National Standards Institute (ANSI) S3 Bioacoustics Working Group 71 has been (slowly) working on a standard that focuses on the calibration (primarily acoustic calibration) of stimuli used to obtain auditory evoked potentials (AEPs). A subset of the working group members met at the American Auditory Society meeting in March 2011. At that meeting, we discussed several issues. We decided to, for the moment, focus on issues relating to AEPs rather than otoacoustic emissions. We discussed technical issues, including safety requirements and tolerances. The discussions benefited by having experts on AEPs as well as manufacturers. The former group understands the needs from a physiological viewpoint, while the latter brings an engineering view of the feasibility, time, and expense of ensuring that devices meet certain specifications.
The issue concerning defining specific reference equivalent threshold sound pressure levels (RETSPLs) for acoustic transients remains challenging. In my own view of this issue, the RETSPLs for clicks and tone bursts that are used in the International Organization for Standardization standard (ISO 389-6) are problematic. The available published data that these RETSPLs are based on are not always obtained at low enough rates to preclude the influence of temporal integration. It is not clear that the audiometric definition of a normal hearing young adult is always sufficiently rigorous in these studies. For tone bursts, the standard only includes RETSPLS for one stimulus envelope, which limits the utility of these data for those using, for example, longer duration tone bursts for long-latency AEPs. Some preliminary estimations of brief-duration RETSPLs for varying envelopes and carrier frequencies, using temporal integration slopes and cutoff durations from the literature, did not produce estimated RETSPLs that consistently related to the RETSPLs specified by the ISO standard. The paucity of published data for very brief duration tone bursts appears problematic, and using this approach to produce estimated RETSPLs for varying tone burst envelopes does not look promising at the present time.
I believe that the use of peak-to-peak peak equivalent SPL (p-p peSPL), as specified in the International Electrotechnical Commission (IEC 60645-3) and ISO (ISO 389-6) standards, is not optimal. For click stimuli, the p-p peSPL can, theoretically, range from 0 dB to 6 dB less than the peSPL obtained by the baseline-peak approach (b-p peSPL). True peak SPL (pSPL) is always 3 dB more than b-p peSPL for a given stimulus but can be anywhere from 3 to 9 dB more than p-p peSPL. I have recently collected acoustic data relating pSPL, b-p peSPL, and p-p peSPL to click stimuli. Supra-aural and insert earphones were used, with a variety of appropriate couplers. The dB peSPL difference between the b-p and the p-p peSPL approaches varied across earphone and coupler. In the draft ANSI AEP standard, we describe how to obtain pSPL, b-p peSPL and p-p peSPL measures of stimulus level. We will likely provide an annex that provides mean peSPL data for the b-p and p-p approaches, for several earphones and couplers. We also will likely provide an annex with the RETSPL data for clicks and tone bursts from the ISO standard. Although annexes are not part of the standard, until we have better threshold data for transient stimuli, these annexes allow end users to use the data in the ISO standard and to convert b-p peSPL or pSPL data to the p-p peSPL values used in the ISO standard.
About the Author
Robert Burkard, PhD, CCC-A, is professor and chair of the department of rehabilitation science at the University at Buffalo, The State University of New York.