Comparing Hearing Aid Algorithm Performance using Simulated Performance Intensity Functions

Simulated performance intensity functions were used to quantitatively discriminate speech intelligibility through phoneme discrimination assessment. Listener test results for subjects with a wide range of sensorineural hearing losses were simulated using an auditory nerve model and compared to real listeners' unaided and aided performance. Simulations of NAL-RP and DSL 4.0 fitting algorithms were compared. Auditory nerve discharge patterns from the model were presented as neurograms. An automated ranking process was used to quantify neurogram degradation using a new measure, the Neurogram Similarity Index Measure (NSIM). The measure has previously been shown to correlate well in predictions of phoneme discrimination for normal hearing listeners in both quiet and noise. In this study, simulated responses to consonant-vowel-consonant word lists in a quiet environment at a range of presentation levels were used to produce phoneme discrimination scores. This represents a further step in validating the use of auditory nerve models to predict speech intelligibility for different hearing aid fitting methods in a simulated environment, allowing the potential for rapid prototyping and early design assessment of new hearing aid algorithms.