Temporal Envelope Coding of the Human Auditory Nerve Inferred from Electrocochleography: Comparison with Envelope Following Responses

Abstract

Neural coding of the slow amplitude fluctuations of sound (i.e., temporal envelope) is thought to be essential for speech understanding; however, such coding by the human auditory nerve is poorly understood. Here, neural coding of the temporal envelope by the human auditory nerve is inferred from measurements of the compound action potential in response to an amplitude modulated carrier (CAP_ENV) for modulation frequencies ranging from 20 to 1000 Hz. The envelope following response (EFR) was measured simultaneously with CAP_ENV from active electrodes placed on the high forehead and tympanic membrane, respectively. Results support the hypothesis that phase locking to higher modulation frequencies (> 80 Hz) will be stronger for CAP_ENV, compared to EFR, consistent with the upper-frequency limits of phase locking for auditory nerve fibers compared to auditory brainstem/cortex neurons. Future work is needed to determine the extent to which (1) CAP_ENV is a useful tool for studying how temporal processing of the auditory nerve is affected by aging, hearing loss, and noise-induced cochlear synaptopathy and (2) CAP_ENV reveals the relationship between auditory nerve temporal processing and perception of the temporal envelope.

Keywords: Auditory nerve; Compound action potential; Electrocochleography; Envelope following response; Human; Temporal envelope.

Fig. 1

Example waveforms measured from tympanic membrane (CAP_ENV, left) and high-forehead (EFR, right) electrodes in response to an AM 3000-Hz carrier in one participant. Each row is a different modulation frequency ranging from 40 to 1000 Hz. Waveform spectra are displayed within the inset above each waveform

Fig. 2

Average CAP_ENV (filled blue circles) and EFR magnitudes (open red circles) for seven participants with normal hearing. Top, middle, and bottom panels show data for the 1000-Hz, 3000-Hz, and NBN carriers, respectively. Noise floors (± one standard error) for CAP_ENV and EFR measures are displayed as blue- and red-shaded regions

Fig. 3

Average CAP_ENV (filled blue diamonds) and EFR synchrony (open red diamonds) for seven participants with normal hearing. Top, middle, and bottom panels show data for the 1000-Hz, 3000-Hz, and NBN carrier, respectively. Noise floors (± one standard error) for CAP_ENV and EFR measures are displayed as blue- and red-shaded regions

Fig. 4

The dB change in magnitude (open circles) or percent change in synchrony (open diamonds) for the tympanic membrane (TM) electrode versus the high-forehead electrode (i.e., TM-benefit) as a function of modulation frequency. The different carrier types are shown as separate panels (top: 1000 Hz carrier, middle: 3000 Hz carrier, and bottom: NBN carrier)

Fig. 5

Comparison of the delta comb analysis (pale orange lines, see text) and the average CAP_ENV (black lines) from seven participants with normal hearing for modulation frequencies from 40 to 1000 Hz (rows). Data are from recordings made with the 3000-Hz carrier, and left and right columns show comparisons of the entire waveform or the first 15 cycles, respectively