Bekesy Audiometry

Introduction :

Bekesy (1947) audiometry  allows the patient to track his own threshold by pressing and releasing a button. The button controls a motor, which in turn controls the attenuator, so that the level increases and decreases at a given rate (usually 2.5 dB/s). The patient is told to hold the button down when he can hear the tone and to release it when he cannot hear the tone. The patient does not press the button whenever the tone is too soft to hear. In this case the motor causes the intensity to rise, so that the tone will eventually become audible. Upon hearing the tone, the patient presses (and holds down) the response button. This causes the motor to reverse so that the intensity decreases. The tone then
becomes inaudible and the patient releases the button, which in turn causes the intensity to rise, and so on. This course of events will cause the level of the tone to rise and fall around the patient’s threshold. The width of the zigzags is often called the excursion width, and the patient’s threshold is the midpoint of these excursions.

Conventional Bekesy Audiometry

Bekesy audiograms are obtained either one frequency at a time or while the test frequency slowly changes from low to high. During sweep-frequency Bekesy audiometry the patient tracks his threshold while the frequency of the test tone increases smoothly from 100 to 10,000 Hz at a rate of one octave/second. During fixed-frequency Bekesy audiometry, the patient tracks his threshold at one frequency for a given period of time, such as 3 minutes. Each Bekesy audiogram is tracked twice, once using a continuous tone and once using a tone that pulses on and off 2.5 times a second, and the results are interpreted by comparing the continuous and pulsed (or interrupted) tracings.

Jerger (1960a, 1962) classified Bekesy audiogramsinto four basic types, although modifications of the original classification system have been described (Owens 1964b; Johnson & House 1964; Hopkinson 1966; Hughes, Winegar, & Crabtree 1967; Ehrlich 1971). The Bekesy types will be described principally in terms of the sweep frequency patterns, examples of which are shown in Fig.1.





Type I 

the pulsed and continuous tone tracings are interwoven, following the pattern of the patient’s audiogram. This type is associated with those who have normal hearing or conductive hearing losses.

Type II

It is associated with cochlear impairments. Here the pulsed and continuous tracings are interwoven for frequencies up to roughly 1000 Hz. Two things happened at higher frequencies. First, the continuous tracing falls below the pulsed tracing by an amount that is usually less than 20 dB, and then runs parallel to it. In addition, the excursions of the continuous tracing narrow considerably, becoming only ~ 3 to 5 dB wide.
The shifting of the continuous tracing in Bekesy audiograms is generally interpreted as revealing the eff ects of tone decay (e.g., Harbert & Young 1964; Owens 1965; Parker & Decker 1971; Silman et al 1981). The narrowed excursion widths in the type II Bekesy audiogram refl ect how intensity perception is aff ected by cochlear impairments, but the mechanism is controversial (e.g., Bekesy 1947; Denes & Naunton 1950; Hirsh et al 1954; Owens 1965). It is probably related to the intensity DL around threshold, but it is not a test of loudness recruitment per sec.

Type III

the continuous tracing diverts very quickly from the pulsed tracing, often shifting to the limits of the audiometer. Type III is associated with retrocochlear pathologies.

 Type IV

the continuous tracing quickly falls more than 20 dB below the pulsed audiogram, and then runs parallel to it. Type IV Bekesy audiograms may occur in patients with cochlear or retrocochlear disorders, but they are often taken to suggest the possibility of the latter (e.g., Turner et al 1984).


Forward-Backward Bekesy Audiometry :

Rose (1962) reported that the diff erence between the continuous and pulsed tracings on sweep-frequency audiograms can be larger when the frequency sweep goes from high to low (backward) compared with the normal direction from low to high (forward). This diff erence is called the forward-backward discrepancy. Large forward-backward discrepancies are associated with retrocochlear pathologies (Kärjä & Palva 1969; Palva, Kärjä, & Palva 1970; Jerger, Jerger, & Mauldin 1972; Jerger & Jerger 1974a; Rose, Kurdziel, Olsen, & Noff singer 1975).
The forward-backward discrepancy using an approach like the one described by Jerger et al (1972) can identify a retrocochlear disorder that would have been missed with standard Bekesy audiometry. There are three Bekesy tracings in each frame, one pulsed and two continuous. The pulsed tracing goes in the forward direction (increasing in frequency from 200 to 8000 Hz). One of the continuous tracings goes in the forward direction (from 200 to 8000 Hz). However, the other continuous tracing goes in the backward direction (decreasing in frequency from 8000 to 200 Hz).
All three tracings are superimposed for the patient’s normal right ear (upper panel). However, there is a forward-backward discrepancy for her left ear, which has an acoustic neuroma (lower panel). Notice that the backward continuous tracing falls far below the pulsed tracing even though there is much less separation between the tracings for the pulsed tone and the forward continuous sweep (which would have been type II).


Bekesy Comfortable Loudness Testing

Bekesy Comfortable Loudness (BCL) testing is similar to conventional sweep-frequency audiometry
except the patient is instructed to press and release the response button to keep the tone “at a comfortable loudness level, neither too loud nor too soft” (Jerger & Jerger 1974a, p. 352). Jerger and Jerger identify three BCL confi gurations considered to be negative because they were associated with normal hearing and with conductive and cochlear impairments. The pulsed and continuous tracings were interwoven in type N1.

Sensitivity and Specificity of Bekesy Audiometry:

On average across studies, the correct identification rates for cochlear and retrocochlear disorders,
respectively, are ~ 93% and 49% for conventional Bekesy audiometry, 95% and 71% for forward-backward Bekesy, and 92% and 85% for Bekesy Comfortable Loudness (Turner et al 1984).