Estimating Uncomfortable Loudness Levels using Evoked Potentials to Auditory Stimuli for Hearing Aid Fitting

摘要

Determining the loudest sound level that a person can comfortably tolerate (uncomfortable loudness level: UCL) imposes a strain on people suffering from hearing loss. In the present study, we propose a method of estimating UCL based on auditory evoked potentials (AEPs). Adults with normal hearing (18 men aged 25-56 years) participated in the study. Three tone bursts (S1, S2 and S3; a triplet) of the same frequency (either 1k, 2k or 4k Hz) were presented to the right or left ear with an interstimulus interval of 300 ms. The sound intensity decreased gradually by 5 dB HL from 80 dB (S1) to 70 dB HL (S3). The interval between triplets was 450 ± 50 ms. The frequency of a given triplet differed from the frequency of the preceding triplet. An electroencephalogram was recorded from three scalp electrode sites (Cz, C3, and C4) with the right mastoid reference. The 900-ms period after the onset of the triplet was transformed to a wavelet coefficient and averaged separately by stimulated ear and tone frequency. The UCLs were estimated by linear discriminant analysis on the basis of trained data of the other participants' subjective UCLs and the wavelet coefficients. The mean estimation error was 4.9 ± 5.0 dB. This result suggests that the UCLs could be estimated successfully on the basis of AEPs to triplets of auditory tones.