Cochlear implants are surgically implanted devices which restore the ability to hear to the hearing impaired. Recent literature has indicated that children receiving cochlear implants (CIs) often have dramatically improved speech and language ability relative to previous generations of children with hearing loss; however, many pediatric CI recipients display persistent speech and language disorders despite early implantation and associated speech/language intervention. Cochlear implants are programmed via mapping - a process in which each individual electrode (FDA approved cochlear implants have between 12 and 22 electrodes) is turned on and the stimulus level adjusted to a level that is comfortable and beneficial to the recipient. At present, this standard of care (SOC) mapping procedure is performed without knowledge of the physical location between the cochlear implant electrodes and the neural interface. Our team has developed a new method of mapping using post-operative CT scans and image processing to specify the physical relationship between the cochlear implant electrodes and the neural interface allowing customized mapping. Using this information, the investigators deactivate sub-optimally positioned electrodes. The investigators term this "Image-guided Cochlear Implant Programming" (IGCIP). This project provides a unique opportunity to examine whether individualized, image-guided CI programming (IGCIP) significantly improves outcomes in pediatric CI patients.

Although children with cochlear implants (CIs) have significantly improved speech, language, and reading outcomes relative to previous generation CI recipients, too many pediatric CI users still display persistent speech, language, and reading difficulties despite early implantation and early intervention. Children with CIs typically lag behind their peers with normal hearing (NH) by 1 or more years on measures of speech, language and/or reading. Though these persistent delays can be attributed in part to a period of auditory deprivation prior to implantation, increasing evidence suggests that a degraded CI signal is also implicated in poorer development of auditory, speech, language, and reading skills for pediatric CI recipients. A related developmental path to reading also disrupted from the degraded CI signal is phonological awareness (PA) because PA is predicated, in part, on speech recognition.

A procedure developed by Noble and colleagues, image-guided CI programming (IGCIP), significantly improves auditory function, speech recognition, and distally, receptive language abilities for adult CI users. The investigators have preliminary evidence that pediatric CI recipients also significantly benefit from IGCIP. But there is a need to systematically investigate IGCIP in children to determine whether this individualized intervention yields a) associated benefits in auditory function and b) related improvements in speech, language, PA and/or reading. Thus, our primary goal is to evaluate the effects of IGCIP on auditory function, speech recognition, PA and reading, as well as speech and language abilities in pediatric CI recipients within the context of a double blind, waitlist controlled randomized clinical trial (RCT). The investigators will obtain psychophysical estimates of auditory function and speech recognition, PA, reading, speech, and language abilities for 72 pediatric CI users in a baseline assessment and repeated time points for 24 months to test the impact of IGCIP. The investigators will examine the immediate (short-term) and longer-term effects over a 2-year period by comparing outcomes between groups for those randomly assigned to immediate (n = 36) or deferred (n = 36) IGCIP using a waitlist control study design (deferred IGCIP). The initial comparison will be for immediate and deferred IGCIP groups at 2, 6, and 12 months. The deferred group will then receive the IGCIP intervention and both groups will be followed for an additional 12 months (total enrollment for 24 months). The proposed research includes the following aims and specific hypotheses:

Aim 1: Auditory function. The investigators will compare auditory function and speech recognition of the immediate and waitlist control participants. Hypothesis 1a: There will be significant positive short-term gains (2-6 months) in spectral and/or temporal resolution as well as speech recognition-particularly in noise-for children immediately receiving IGCIP as compared to waitlist controls. This hypothesis will be tested by comparing the difference in the amount of change in scores within-subjects (pre- to post-IGCIP gain) between the groups (treated vs. untreated) controlling for initial scores. Hypothesis 1b: IGCIP gain in spectral and/or temporal resolution will significantly predict gain in speech recognition. This hypothesis will be tested via regression analyses of change in speech recognition scores on change in resolution, controlling for baseline values and also controlling for baseline levels of speech recognition and working memory.

Aim 2: PA and reading. The investigators will explore the complex relationships amongst auditory function, speech recognition, PA, and reading ability. Hypothesis 2a: Differential growth in spectral/temporal resolution and/or speech recognition will predict growth in PA, which in turn will predict mediated growth in reading. Hypothesis 2b: Growth in PA will be associated with amount of IGCIP benefit (gain) and will mediate growth in reading, which will be tested via cross-legged panel and path analyses. Note that testing these hypotheses is not dependent on the outcomes of Aim 1 as only variable gain in the Aim 1 measures (e.g., speech recognition) are required for aim 2 analyses, not a significant between-group difference for IGCIP in Aim 1.

Aim 3: Speech and language. The investigators will compare pre- and post-IGCIP receptive and expressive language abilities and speech production of pediatric CI recipients to the waitlist control group. The investigators will test these skills at various time points on standardized and clinical measures of 1) receptive language, 2) expressive language, and 3) speech production (articulation and acoustic analyses). Hypothesis 3a: There will be significant differences between groups for positive growth in speech and language and this growth will be predicted by the relative improvement in auditory function (aim 1) from IGCIP while controlling for baseline levels of working memory. Hypothesis 3b: Spectral/temporal resolution and speech recognition and/or PA will serve as mediators of expressive and receptive language gains and speech production gains both within and between groups. 3a and 3b will also be tested using mixed effects modeling and regression analyses to examine these "downstream" effects. Even if no between group differences in Aim 1 and/or Aim 2 are seen, the investigators will nonetheless be able to test whether spectral/temporal resolution, speech recognition, and/or PA predict growth in receptive and/or expressive language and/or changes in speech production (including subclinical acoustic analyses).

Inclusion Criteria:

1. Children aged 4 to 12 years of age

2. Prelingual onset of deafness

3. At least one CI and bilateral moderate to profound sensorineural hearing loss

   • for children with a single CI, audiometric thresholds in the non-CI ear must be consistent with at least a moderate to profound sensorineural hearing loss

4. Cochlear implantation prior to 4 years of age

5. Nonverbal cognitive abilities within the typical range

6. No confounding diagnosis such as autism spectrum disorder, neurological disorder, or general cognitive impairment

7. Pre-operative CT scan of head performed as standard of care CI work-up

8. Post-operative CT scan--obtained either before enrollment (per VUMC CI program standard of care) or after informed consent

Exclusion Criteria:

1. Severe anatomical abnormality(s) of the temporal bone.
2. Onset of moderate-to-profound sensorineural hearing loss after 2 years of age
3. Nonverbal intelligence standard score < 85

A randomization schedule will be generated by study statistician and provided to PIs prior to study commencement. To ensure equal numbers of participants in each arm, a computer-generated, permuted blocking algorithm (blocks of 4 participants) will be used to develop the schedule. The schedule will be password protected and saved on an encrypted server housed at the Vanderbilt Bill Wilkerson Center. The investigators will be using identical procedures for all participants regardless of arm to which the participants randomize. Both the experimenters and the participants will be blinded. The experimenters will be notified of the randomization for a given participant on the day of the baseline visit. Only the PIs and study statistician will know whether the participant is in the intervention or deferred waitlist group until the end of the study. Neither PI nor study statistician will be personally administering assessments nor scoring tests for the participants.