Comparison of clinical feasibility of behavioural and physiological estimates of peripheral compression

Previous research has shown that the rate of peripheral compression (estimated from the slope of the basilar-membrane input-output function) is not correlated with the pure-tone sensitivity (the audiogram). However, efficient estimation of peripheral compression has proven challenging and the methods are based on several assumptions. The aim of this study was to investigate and compare results from three methods of estimating peripheral compression in terms of their accuracy and clinical feasibility. Two psychoacoustic behavioural measures, based on forward (temporal masking curves, TMC) and simultaneous masking with notched-noise (NN), were investigated together with a physiological, distortion-product otoacousticemissions (DPOAE), based measure. Forty-five hearing-impaired (HI) listeners with mild-to-moderate hearing loss were tested. Correlation analysis of the data was performed, including partial-correlations, in order to factor out the potential influence of the pure tone-thresholds on the compression estimates. The results demonstrated limitations of each of the considered methods; however, the experiment involving estimates of auditory filters showed good stability and small training requirements across the listeners.