The influence of signal level and temporary noise induced hearing loss on estimates of the post-stimulus time histogram and single fiber action potential derived from human compound action potentials

An analytic compound action potential (CAP) designed by convolving functions representing the post-stimulus time histogram summed across auditory nerve fibers [P(t)] and a single fiber action potential [U(t)] was fit to human CAPs. Applying the analytic CAP to click-evoked CAPs before and after a noise-induced temporary hearing threshold shift (TTS) provided estimates of the number of neurons contributing to the CAPs (N), as well as in vivo P(t) and U(t), in normal and temporarily damaged cochleae. Click levels varied from 75 to 125 dBpSPL in 10 dB steps. Hierarchical Linear Modeling was used to characterize changes in the parameters governing the profile of N, P(t) and U(t) as a function of signal level. The width of P(t) decreased with increasing signal level. During TTS, P(t) width increased at the lowest signal level indicating less synchrony of neuronal discharge. The latency of P(t) decreased with increasing signal level. Following noise exposure, P(t) latency was shorter at all signal levels suggesting that neurons of more basal cochlear location dominated the response following noise exposure. Parameter N increased with signal level. During TTS, the lowest signal level N decreased indicating the number of responding neurons was reduced. U(t) rang less with increasing signal level. For subjects with large amounts of TTS, low signal level U(t) rang less. The frequency of U(t) oscillation increased with signal level. Following noise exposure, U(t) oscillated with a lower frequency at all signal intensities. Together, these results show that during TTS less neurons contribute to the CAP and these neurons have shorter latency and less response synchrony. Moreover, the oscillation of single fiber action potentials decay more rapidly and with a lower frequency during TTS.