noise, and thus absent OAEs in low-frequency regions should not be given great weight in interpretation.
OAE presence indicates good hair cell function and generally indicates that the hearing thresholds should be better than 30-40 dB. OAEs, however, cannot be used to determine exact hearing thresholds. In contrast, the absence of an OAE can be due to a variety of causes, from middle ear dysfunction to sensorineural disorders producing hearing loss of any degree. The absence of an OAE alone should not be interpreted as indication of significant hearing loss. There are conditions in which the presence of an OAE alone does not ensure normal hearing sensitivity. Disease that spares the cochlea and impairs function in the auditory nerve or low brainstem (for example acoustic neuroma or auditory neuropathy) can also cause significant hearing loss. The OAE therefore should not be tested in isolation but must be included in a battery of tests for accurate interpretation. Nevertheless, measurement of OAEs provides a quick, noninvasive view of the functioning of the inner ear. Because of the value of these measures, the evaluation of OAEs is routine in many diagnostic audiology settings.
Auditory evoked potentials (AEPs) are recordings of neural activity evoked by sound. The AEP is collected using surface electrodes placed on the scalp and near the ear. Computer-generated sounds are presented to subjects via earphones, and each presentation triggers a synchronized recording of neural activity. The responses to many stimuli are averaged in a manner time-locked to the stimulus to reduce the contribution of nonauditory generators, such as random muscle or brain activity.
AEPs occur within fractions of a second following stimulation. The earliest activity, known as the auditory brainstem response or ABR, has a latency of 0-20 ms depending on the age of the subject and the nature of the sounds used to elicit the response. This response is generated in the auditory nerve and brainstem auditory pathway. Other evoked potentials include the middle latency response (MLR), generated in the thalamus and primary auditory cortex, with a latency of 20-100 ms, and the late cortical response (LCR), generated in the auditory cortex and association areas, with a latency of 100-250 ms. While all of these AEPs can be used to predict hearing threshold levels, the one used most commonly in the United States is the ABR. In contrast, most of the audiology literature from other parts of the world supports the use of late responses for threshold estimation in adults (Coles and Mason, 1984; Hone, Norman, Keogh, and Kelly, 2003; Hyde, Alberti, Matsumoto, and Yao-Li, 1986; Tsui, Wong, and Wong, 2002). Most recently, a variation of standard evoked potential