Table 1.
Participant demographics.
Table 2.
Auditory brainstem recording parameters.
Table 3.
A) Stimulation current levels (dB re: 100 µA) in the group of adolescent cochlear implant users, B) stimulation current levels (dB re: 100 µA) in the experienced bilateral cochlear implant users implanted sequentially or simultaneously, C) stimulation levels in dB SL (Sensation Level, re: behavioral threshold) in the group with normal hearing.
Figure 1.
A)Thirty four adolescents responded to bilateral input either with no intended interaural level differences (ILD = 0) as indicated by equal amplitudes of auditory brainstem responses or with ILDs of 10 or 20 weighted to the second CI (positive values) or first CI (negative values) at early stages of bilateral implant use (0.14±0.01 year). Unilateral stimuli (CI1 and CI2) were also presented randomly to ensure the task was understood. Upper plot: Mean (±1 SE) proportion of responses from all participants. Mean responses to unilateral stimuli were>0.88 accurate. Significant changes in mean proportion of responses occurred with changing ILD (CI1 responses: p<0.0001, CI2 responses: p<0.001). ILDs of 20 CU (+ values indicate CI2 weighted and – values indicate CI1 weighted) were perceived to come from the correct side in 0.75 of trials with decreasing certainty for ILDs of 10 CU ILD and no significant difference between proportion of CI1 and CI2 responses when no ILDs were provided (ILD = 0) (p>0.05). Lower plot: Logit regression analyses of the proportion of right responses from each child revealed significant changes in 21 participants (solid lines) and non-significant changes in 9 participants (dashed lines). Data from 4 children was excluded from analysis because <0.67 of responses to unilaterally presented stimuli were accurate. White diamonds represent mean (±1 SE) data from significant regression curves. B) ILD perception was retested after 0.86±0.03 year of bilateral use in 29 of the 34 children. The “middle” and “both” choices were not accepted at this test time. Upper plot: significant change in mean responses were found with changing ILD (CI1 and CI2 p<0.0001). Lower plot: 25 of 29 adolescents showed significant changes in proportion of CI1 responses with ILD (solid lines), 3 had insignificant changes in responses (dashed lines), and data from 1 participant was excluded because responses to unilaterally presented stimuli were accurate on <0.67 of trials. White diamonds represent mean (±1 SE) data from significant regression curves.
Figure 2.
A) Mean (±1 SE) proportion of responses from all participants in three groups. Positive interaural level differences (ILDs) are left weighted and negative values are right weighted. CI groups had long term bilateral implant experience. Significant effects of ILD were found across groups (p<0.0001) with no effect of group (p>0.05) and no interaction (p>0.05). B) Logit regression curves for proportion of right responses from each participant for each group. Significant changes were found for all participants (solid lines) bar 4 of 29 children in the simultaneous group (dashed lines). Slopes were significantly reduced in the implanted groups relative to the Normal Hearing Group (p<0.0001).
Figure 3.
Experienced bilateral CI users and normal hearing peers perceived changes in ILDs as these cues moved from left to right weighted.
Interaural level differences (ILDs) in CI users represent differences in CU and dB re: 20 Pa in the normal hearing group (∼0.08 dB re: 100 µA per acoustic dB change in ILD, see Table 3 for additional details). Significant effects of ILD were found across groups (p<0.0001). Rate of change in ILDs delivered by CIs (CU) was reduced relative to acoustic ILD (dB) (p<0.05) with no significant difference between the 2 implanted groups (p>0.05).
Figure 4.
A) Upper plot: Mean (±1 SE) proportion of responses to changes in interaural timing differences (ITDs) at early stages of bilateral implant use (0.14±0.01 year). Positive values lead from the newly implanted ear (CI2) and negative values from the more experienced ear (CI1). Most responses are to either CI1 or CI2. No significant changes were found for CI1 or CI2 responses with ITD changes (p<0.05). Lower plot: Logit regression curves for proportion of right responses from each participant reveal non-significant slopes in 15 participants (dashed lines) and significant slopes (solid lines) in 4 participants but in the opposite direction from expected. Data could not be analyzed in 2 children as all responses were to CI2, 3 children were excluded from individual because of non-significant changes in responses to interaural level differences (shown in Figure 1A), and 1 was excluded because responses to unilateral stimuli were <67% accurate. B). Upper plot: No significant changes in CI1 or CI2 responses were found after 0.86±0.03 year of bilateral use. Lower plot: Logit regression curves showed non-significant changes in 17 participants (dashed lines) and significant changes in 6 participants (solid lines) although 1 was in the opposite direction from expected. Data could not be analyzed in 2 adolescents who indicated CI2 for all responses, 3 participants with non-significant changes with interaural level differences (Figure 1B) and 1 whose responses to unilaterally presented stimuli were <67% accurate.
Figure 5.
A) Mean (±1 SE) proportion of responses from all participants in three groups. Positive interaural timing differences (ITDs) indicate bilateral stimuli leading from the left and negative values indicate right leading stimuli. CI groups had long term bilateral implant experience (Simultaneous Group: 4.31±1.10 years; Sequential Group: 5.30±1.55 years). Mean responses from the Normal Hearing Group are essentially at ceiling for all ITDs>0. B) Logit regression analyses of data from individual children are shown (solid lines = significant regression curves; dashed lines = non-significant regression curves). Twenty of 25 in the Simultaneous Group significantly detected changes in ITD (data from 5 were not analyzed as detection of interaural level cues shown in Figure 2B was not significant). Eleven of 16 children in the Sequential Group showed significant changes. There was no significant difference between these proportions (Mann-Whitney U = 177.5, p = 0.42). Significant changes were found in all children in the Normal Hearing Group.
Figure 6.
Logit regression curve analyses from Figures 3B and 5B were used to calculate predicted balance (0.5 proportion of right responses) for both interaural level and timing differences (ILDs and ITDs), revealing time-intensity trading in all groups.
Predictions of negative ILDs for balanced bilateral input indicated that bilateral stimuli at ILD = 0 was left weighted and predictions of positive ILDs for balanced bilateral input indicated ILDs = 0 was right weighted. The perception of ITDs to stimuli presented at ILD = 0 shifted predictably in each group; linear regressions were significant in both CI groups (p<0.0001) with a trend in the Normal Hearing Group (p = 0.054). Left weighted stimuli shifted balanced perception toward right leading stimuli and right weighted stimuli shifted balanced perception toward left leading stimuli.
Figure 7.
Mean (±1 SE) proportion of right responses from children with significant changes in interaural timing differences (ITDs) from the Normal Hearing Group (n = 16), the Sequential CI Group (10 with CI1 in the right ear), and the Simultaneous CI Group (n = 20).
Responses changed significantly with ITD (p<0.0001) with no group effect (p>0.05) but with a group*ITD interaction (p<0.0001). The proportion of right responses for stimuli left leading by 400 µs was significantly greater in the Sequential CI group (p<0.005). The rate of change was significantly reduced in the CI groups relative to normal (p<0.0001) and there was a trend toward reduced rates of change in the Sequential CI group relative to the Simultaneous CI group (p = 0.05).