24.10.2012
Determining the threshold of hearing with pure-tone stimuli is the most widely used and most fundamental audiometric procedure. As for all measurement procedures, correct calibration plays a decisive role for the reliability of the results obtained. In the new version of EN ISO 8253-1 (Acoustics – Audiometric test methods – Part 1: Basic pure-tone air and bone conduction threshold audiometry), additional statements are made in order to draw up a measurement uncertainty budget complying with the requirements of the "Guide to the expression of uncertainty in measurement" (GUM). Our research work has shown that the interindividual differences in the geometry of the ear canal lead to an uncertainty of the hearing threshold determination that increases along with frequency. Relating the individual hearing threshold to the individual eardrum sound pressure (instead of using mean reference hearing thresholds) would allow the hearing threshold to be determined more accurately.
So-called "Reference Equivalent Threshold Sound Pressure Levels" (RETSPL), which are defined in the EN ISO 389 series, are required to calibrate an audiometer and a given sound transducer (such as, e.g., the Sennheiser HDA 200200, and Telephonics TDH-39, headphones, the insert earphone Etyomitc Research ER-3A insert earphone or for free-field or diffuse-field presentation). These equivalent reference threshold sound pressure levels describe the mean hearing threshold (determined under laboratory conditions) of a group of otologically normal persons aged 18-25. EN ISO 8253-1 specifies both the requirements for the measuring instruments and the measurement environment, and the audiometric measurement procedure to be used. In addition, a total of eight quantities that have an influence on the total uncertainty of the hearing threshold when applying pure-tone audiometry are listed in Annex A of EN ISO 8253-1. As far as the hearing threshold level for airborne sounds is concerned, the four most important of them are::
- L'HT - the hearing threshold level determined. If the measurement is repeated under identical conditions, based on experience, a standard uncertainty of 2.5 dB (up to 4 kHz) and 4 dB (from 4 kHz) can be expected.
- δeq - Deviations of the audiometric instruments used from the reference behaviour. For instruments of class 1 and 2, according to IEC 60645-1 (Electroacoustics - Audiological equipment Part 1 – Pure-tone audiometers), a standard uncertainty of 1.7 dB (up to 4 kHz) and 2.3 dB (from 4 kHz), respectively, is assumed. These uncertainties increase to 2.3 dB and 2.7 dB, respectively, when the increment of the hearing level control is 5 dB. Furthermore, one assumes that there is no systematic error, i.e. the expected value for δeq is 0 dB.
- δtr – Deviations due to RETSPL values that are not fully equivalent in ISO 389, as well as deviations caused by the fitting of the earphone and individual anatomic and physiological characteristics of the test subject. In this regard, a standard uncertainty of 2.9 dB (up to 4 kHz) and 3.9 dB (from 4 kHz) is assumed. Also in this case, one assumes that there is no systematic error.
- δn – Influence of background noise. For test subjects with very good hearing, a standard uncertainty of 2 dB is assumed. In the case of test subjects having an elevated hearing threshold or in the event of very quiet measurement environments, δn may be neglected.
Compared with variations of the hearing threshold of 20 young, otologically normal persons measured by Schmidt et al. (2011) [1], this estimate provides a satisfactory agreement. In fact, the experimentally observed variation at frequencies below 4 kHz is slightly higher, and that observed at frequencies between 4 kHz and 9 kHz slightly lower than the estimate stated in EN ISO 8253-1 (see Figure 1). At frequencies of approx. 6 kHz and higher, the individual anatomic and physiological characteristics of the test subject contribute considerably to the measurement uncertainty. This uncertainty contribution could be reduced if it were possible to reliably determine the hearing threshold as a function of the sound pressure at the individual eardrum, see Schmidt et al. (2011) [1]. Due to the high vulnerability of the ear canal and of the eardrum, measuring directly at the eardrum is impracticable. Measuring the sound pressure at the entrance of the ear canal and estimating the sound pressure at the eardrum applying certain models, however, seems to make more sense but is, to this day, still object of research.
Fig. 1: Estimated interindividual standard deviation (dashed) for the hearing threshold levels, obtained by means of an automatic-recording sweep-frequency audiometer, of 20 left ears (dashed grey) and 20 right ears (dashed black). The combined uncertainty according to Annex A of EN ISO 8253-1 is depicted as a solid black line.
References:
[1] Schmidt J.-H., Mauermann M., Blau M. (2011): "Streuung der Hörschwelle von Normalhörenden für Frequenzen oberhalb 1 kHz bei Bezug auf den Schalldruck im Ohrsimulator, im Freifeld und am Trommelfell". DAGA 2011 Düsseldorf (in German).
Contact persons:
Jan-Henning Schmidt, Dept. 1.7, WG 1.72, E-Mail: jan-henning.schmidt@ptb.de
Thomas Fedtke, Dept.FB 1.6, WG 1.61, E-Mail: thomas.fedtke@ptb.de