Exploration of stimulus-frequency otoacoustic emission suppression tuning in hearing-impaired listeners

Abstract

Objective: Otoacoustic emissions (OAEs) can provide useful measures of tuning of auditory filters. We previously established that stimulus-frequency (SF) OAE suppression tuning curves (STCs) reflect major features of behavioral tuning (psychophysical tuning curves, PTCs) in normally-hearing listeners. Here, we aim to evaluate whether SFOAE STCs reflect changes in PTC tuning in cases of abnormal hearing.

Design: PTCs and SFOAE STCs were obtained at 1 kHz and/or 4 kHz probe frequencies. For exploratory purposes, we collected SFOAEs measured across a wide frequency range and contrasted them to commonly measured distortion product (DP) OAEs.

Study sample: Thirteen listeners with varying degrees of sensorineural hearing loss.

Results: Except for a few listeners with the most hearing loss, the listeners had normal/nearly normal PTCs. However, attempts to record SFOAE STCs in hearing-impaired listeners were challenging and sometimes unsuccessful due to the high level of noise at the SFOAE frequency, which is not a factor for DPOAEs. In cases of successful measurements of SFOAE STCs there was a large variability in agreement between SFOAE STC and PTC tuning.

Conclusions: These results indicate that SFOAE STCs cannot substitute for PTCs in cases of abnormal hearing, at least with the paradigm adopted in this study.

Keywords: Frequency selectivity; hearing impairment; otoacoustic emissions; psychophysical tuning curve; suppression tuning curve.

Figure 1

Hearing thresholds normalized to mean hearing thresholds of normal-hearing young adults, so that the 0 line corresponds to thresholds of 22–35 year old adults (from Lee et al, 2012). Based on the relative hearing threshold level at 1 kHz or 4 kHz, each subject was qualified to one of three experimental groups: <10 dB HL (light grey box), 10–25 dB HL (medium grey box), or 25–45 dB HL (dark grey box). The number of listeners assigned to each group at 1 kHz and 4 kHz is shown in the appropriate shade of grey within the figure. The audiograms of three listeners (kc32, 36, and 39) for whom additional OAE data are displayed in Figure 6 are emphasized with dashed lines and symbols with the listeners’ IDs.

Figure 2

The widths of PTCs (bandwidth 10 dB above the tip in kHz, BW10) as a function of probe level (dB SPL) at 1 kHz (A) and 4 kHz (B). Each data point (marked as listener’s ID number) represents the mean width (BW10) for upward- and downward-sweep PTCs for each listener (note: at 4 kHz BW10 could be calculated only for the upward-sweep PTC for kc28 and only for the downward-sweep PTC for kc39). For listener kc27 at the 1 kHz probe frequency and for kc36 at 4 kHz the PTCs were too “flat” to calculate BW10s (Figure 3, C, D). For comparison, the BW10s for PTCs collected in listeners with normal hearing are shown with black triangles (Charaziak et al, 2013). All PTCs were measured at 10 dB SL with the exception of two listeners who were also tested at 20 dB SL and 30 dB SL (dotted lines). The HL group assignments are shown with grey-shaded boxes.

Figure 3

Individual LOESS-smoothened PTCs (upward masker sweep) at 1 kHz (A and C) and 4 kHz (B and D). Data for the 10–25 dB HL group are shown in panels A and B, and for the 25–45 dB HL group are shown in panels C and D. The saturation of the grey shading corresponds to the degree of hearing loss within a given HL group, with darker shading indicating higher thresholds (see Table 1, Table 2, and listeners’ ID number listed near each curve). For comparison an average PTC for normal-hearing listeners is shown in each panel (black dashed line; Charaziak et al, 2013).

Figure 4

Median SFOAE input-output functions (solid black) and median noise levels (dashed grey) for probe frequencies 1 kHz (A–C) and 4 kHz (D) across different HL groups. The error bars represent 25th and 75th percentiles.

Figure 5

Examples of SFOAE STCs obtained for the 10–25 dB HL listeners (A, B) and the 25–45 dB listeners (C, D) at 1 kHz (A, C) and at 4 kHz (B, D). The more saturated the grey line the higher the SFOAE criterion that was used while the line type (dotted, solid, and dashed-dotted) codes the probe level (see the legend; residual criterion in dB SPL/probe level in dB SL). The PTCs for upward- and downward-sweeps are shown in black dashed lines. For listener kc42 the BW10 ratios ranged from 0.9 to 2.2. For listener kc37 (B) meaningful BW10 could be calculated only for 0/20 SFOAE STCs, resulting in a BW10 ratio of 0.65. For listener kc28 the BW10 ratio was equal to 1.6. For listener kc32 (D) the BW10 ratios were 1.5 and 1.6 (excluding 6/20 STC for which it was not possible to calculate BW10).

Figure 6

SFOAE (A, C, E) and DPOAE (B, D, F) levels as a function of the probe frequency (f₂ for DPOAEs) for three listeners (rows).