Traditionally, psychophysical data have been predicted either by constructing models of the peripheral auditory system or by applying signal detection theory (SDT). Frequently, the theoretical detection performance predicted by SDT is greater than that observed experimentally and a nonphysiologically based "internal noise" source is often added to the system to compensate for the discrepancy. A more appropriate explanation may be that traditional SDT approaches either incorporate little or no physiology or make simplifying assumptions regarding the density functions describing the physiological data. In the work presented here, an integrated approach, which combines SDT and a physiologically based model of the human auditory system, is proposed as an alternate method of quantifying detection performance. To validate this approach, the predicted detection performance for a simultaneous masking task is compared to predictions obtained from traditional methods and to experimental data. Additionally, the sensitivity of the integrated method is thoroughly investigated. The results suggest that by combining SDT with a physiologically based auditory model, thereby capitalizing on the strengths of each individual method, the previously observed discrepancies can be partially explained as the result of physical processes inherent in the auditory system rather than unspecified "internal noise" and more accurate predictions of psychophysical behavior can be obtained.

No abstract

Psychoacoustic experiments indicate that the human auditory system is less sensitive than predicted by classical signal detection theory on a simultaneous masking task [C. M. Reed et al., J. Acoust. Soc. Am. 53, 1039–1044 (1973)]. Traditionally, the difference between the matched filter and the experimental performance has been compensated for by assuming an additional internal additive noise. However, this noise has not been completely explained in physiological terms. The inconsistency may also partially be the result of the suboptimal application of signal detection theory. To investigate this discrepancy, signal detection theory is integrated with Patterson’s model of the human auditory system [R. D. Patterson et al., J. Acoust. Soc. Am. 98, 1890–1895 (1995)]. The performance of the optimal detector for a simultaneous masking task is compared to experimental data, allowing the theoretical performance bounds to be determined and the model to be verified. The results of this work suggest that with a more appropriate application of signal detection theory it is possible to partially explain the difference in performance between the matched filter and experimental data as the result of physical processes inherent in the auditory system. Utilizing a priori information regarding these processes results in an approach which more accurately predicts psychophysical behavior.

Traditionally, psychophysical data has been predicted either by constructing models of the peripheral auditory system or by using signal detection theory. The resulting theoretical detection performance is often substantially greater than experimental results and this discrepancy has been explained by adding '[internal noise" to the system. An integrated approach combining both a physiologically-based auditory model and clwsical signal detection theory has been shown to produce more accurate predictions of experimental data (1). The work presented here investigates the impact of ph~e uncertainty and multiple channel integration on predicted detection performance.

Historically, signa[ detection theory (SDT) applied "outside" the ear has not provided valid estimates of the performance measured in psychophysical experiments. Our previous work, in which SDT wm applied at the output of each stage of the Auditory Image Model (AIM) (1), has indicated that this integrated approach affords theoretical performance predictions that are more closely matched to experimental data than previous approaches (2). In this paper, we describe the results obtained when a similar approach is applied to the output of the Camey model of auditory processing (3). Our results indicate that for a simultaneous masking task, when it is assumed that all parameters of the input signal are known, the performance predictions obtained from SDT integrated with AIM and the Carney model are similar at the neural firing stage, and both over-predict experimental measures. However, when it is assumed that the phase of the sinusoid is unknown, the integrated approach provides performance estimates that are fairly close to the performance levels observed experimentally.