Air-radiated sound from bone transducers

Audiologist need to be aware that many common bone oscillators used in bone conduction measurements can introduce measurement variability through air-borne radiation.   

According to ANSI S3.6 (1996), the manufacturer of the bone device must state the test frequencies at which the bone vibrator can radiate air-conducted sounds which are high enough to invalidate the measurement of bone-conduction thresholds in an unoccluded ear. 

The equipment used in bone conduction measurements is more susceptible to external factors than air conduction methods, and this may affect measurement accuracy of the bone conduction hearing threshold level (Shipton et al 1980).  When performing bone conduction audiometry it is assumed that the vibrations from the bone oscillator stimulate the cochlea through the bone only.   This however many not be the case.  The bone oscillator may set the surrounding air into vibratory motion and this disturbance in the air may enter the ear canal as air borne radiation.  If this happens, the signals presented through the bone oscillator will be perceived as subjectively louder than the vibration-induced sensation, which may result in a false bone conduction threshold and also a false air-bone gap.     

According to International Standards IEC 645 (1979), the air radiation output of a bone vibrator should be at least 10 dB below the vibration output.  For example, if you were testing by bone conduction at 40 dB HL at 1000 Hz, the air borne radiation from the bone oscillator must not exceed 30 dB according to IEC standards.  ANSI S3.6 (1969) required that the air radiation be at least 5 dB below the vibration output, however, this standard has since been replaced by ANSI S3.6 (1996) and this requirement is no longer in place.  The requirement now is that manufacturers indicate the frequencies and levels which air borne sound will invalidate threshold measures.  Most Audiologists will most likely not have to physically test for air borne radiation, but those interested in knowing more about how to test for this are recommended to read section 5.4.3 of ANSI S3.6 1996.   

A study by Shipton et al. (1980) investigated air borne radiation for several commonly used bone oscillators, which included the Radioear B71, B72 and B70A.   Shipton et al. (1980) discovered that air borne radiation is highly directional when measured on an artificial mastoid.  Therefore, depending on the orientation of the oscillator, the air borne radiation can differ by up to 15 dB.  Since these results were obtained on an artificial mastoid rather than an actual person, it was considered to be an unreliable quantitative predictor of the likely effects of air borne radiation in actual test situations.  However, the results from the artificial mastoid can be useful for comparing air borne radiation across different bone oscillators.     

Shipton et al. (1980) also looked at the actual SPL entering the ear canal when the bone vibrator was placed on a human head.  The graph below illustrates the “Margin of safety” demonstrated by each of three common bone oscillators at frequencies normally tested using bone conduction.  

(taken from Shipton et al 1980)

On the Y-axis, the margin of safety at 10 dB represents the level at which IEC standard 645 requires air borne radiation from bone oscillators to fall on or below.  The margin of safety at 5 dB represents the requirements set by ANSI 3.6 (1969).  Where the curve penetrates the line at 0 dB, air borne radiated sound from the bone oscillator will be heard rather than the bone oscillator itself. 

The Shipton et al. (1980) study has found that at 250 Hz, the Radioear B70A produces the highest level of air borne radiation, whereas the B72 produces the lowest.  From 500 to 2000 Hz, the average level of air borne radiation across all three bone oscillators is at least 7 dB below the normal threshold of hearing (meets ANSI 3.6 but not IEC 645).  At 3000 and 4000 Hz, the level of air borne radiation is too high for an accurate measurement of bone conduction hearing thresholds according to Shipton et al (1980).  It is important to note that the SPL that enters the ear canal is only a useful measure when it is compared to a subject’s threshold.  Only then is it possible to determine if the air borne radiation is at a low enough level for accurate bone conduction threshold measurements.  In a case where a subject presents with air conducted hearing thresholds that are within normal limits, the impact of bone conduction accuracy 3 and 4 kHz (or for any frequency) may not be clinical significant.  However, the situation at 3 and 4 kHz does become of interest when a loss of hearing exceeds normal limits by air conduction.

Audiologists need to be aware that the values given in the Shipton et al (1980) study are averaged values, and variation above or below can be expected in the calculation of an air bone gap for different subjects.    

The solution to the problem of air borne radiation at 3 and 4 kHz can be easily overcome by occluding the ear canal with an ear plug during testing of these frequencies.  By inserting an earplug on the side of the ear where the bone oscillator is positioned, air borne radiation can be blocked from entering the canal.  The major drawback to using earplugs is that at low frequency bone conduction testing, earplugs cannot be used.  Blockage of the ear canal at frequencies below 1000 Hz will result in an occlusion effect for that ear.  Having earplugs out for frequencies below 1000 Hz and then inserting them for testing at 3000 and 4000 Hz can be a major inconvenience for Audiologists, especially in the private practice setting where time is more limited.  For this reason, ear plugs in bone conduction testing is often not used.        

The results of the Shipton et al. (1980) study indicates that the B72 is the poorest performing bone oscillator when it comes to air borne radiation.  However, Audiologists should know that it is unlikely that large errors will result from using either one of the three bone oscillators if ear plugs are inserted onto the test ear at frequencies above 3 kHz.

Note:  This site is best viewed using IE 4.X or higher under 800x600 resolution