Speech Intelligibility and Spectral Resolution Results in Cochlear Implant Users With and Without Inner Ear...

This study aimed to compare speech perception in noise and spectral resolution between CI users with normal cochlea and those with inner ear malformations (IEMs) and to examine the relationship between angular insertion depth (AID).

Prospective cohort.

Cochlear implant (CI) program at a tertiary medical center.

Thirty-seven pediatric CI users participated: 14 with normal cochleae and 23 with IEMs (2 with IP-I, 15 with IP-II, 1 with IP-III, and 5 with CH-III). The normal cochleae group consisted of 3 unilateral CI users, 7 bimodal users, and 4 bilateral users, whereas the IEM group consisted of 6 unilateral CI users, 10 bimodal users, and 7 bilateral users. All participants received MED-EL devices. A total of 48 implanted ears from 37 patients were included. In the IEM group, 18 ears (1 with IP-I, 1 with CH-III, and 16 with IP-II) received the FORM24 electrode, and 10 ears (1 with IP-I, 1 with IP-II, 2 with IP-III, and 6 with CH-III) received the FORM19 electrode.

Speech perception was assessed using the hearing in noise test for children (HINT-C), and spectral resolution was measured using the Spectral-Temporally Modulated Ripple Test (SMRT). AID was determined using postoperative transorbital x-rays. Associations between AID and auditory outcomes were examined using correlation and linear mixed-effects modeling.

CI users with normal cochleae had better speech perception signal-to-noise ratio (SNR) in the NIpsilateralSFront and NFrontSFront conditions, whereas CI users with IEMs demonstrated poorer speech perception in noise, requiring more favorable SNRs. There was no significant difference in the NContralateralSFront condition. CI users with IEMs had lower ripple-density thresholds, indicating reduced spectral resolution. In simple regression analyses, deeper AID was associated with better performance in the Quiet and NIpsilateralSFront conditions and with higher SMRT scores. However, in linear mixed-effects models, group differences remained significant for Quiet, NIpsilateralSFront and NFrontSFront, whereas AID independently influenced only Quiet and NIpsilateralSFront. No independent effect of AID on SMRT was observed.

CI users with IEMs required higher signal-to-noise ratios to understand speech in noise and demonstrated poorer spectral resolution than their CI peers with normal cochleae. Although deeper AID was positively correlated with both spectral-resolution skills and speech perception, its independent contribution to spectral resolution appears limited after adjusting for clinical and demographic factors. These results suggest that cochlear anatomy and electrode insertions have potential clinical roles in auditory outcomes, highlighting the multifactorial nature of performance variability in pediatric CI users.