Round Window Electrocochleography to Low Frequency Tones in Pediatric Cochlear Implant Recipients with and Without Auditory Neuropathy Spectrum Disorder: Separating Hair Cell and Neural Contributions Using a Computational Model. Issue 7 (29th August 2022)

• Record Type: Journal Article
• Authors: Fontenot, Tatyana E.; Giardina, Christopher K.; Brown, Kevin D.; Fitzpatrick, Douglas C.
• Is Part Of: Otology & neurotology. Volume 43:Issue 7(2022)
• Page Start: 781
• Page End: 788
• Publication Date: 2022-08-29
• Subjects: Auditory nerve neurophonic -- Auditory neuropathy/dyssynchrony -- Cochlear microphonic -- Computational modeling -- Hair cell -- Sensorineural hearing loss -- Signal analysis; Otology -- Periodicals; Ear -- Diseases -- Periodicals; Skull base -- Surgery -- Periodicals; 617.8005
• DOI: 10.1097/MAO.0000000000003568
• Languages: English
• ISSNs: 1531-7129

Abstract

Abstract : Hypothesis: Characterize the contribution of the auditory nerve neurophonic (ANN) to electrocochleography (ECochG) of pediatric cochlear implant (CI) recipients with and without auditory nerve spectrum disorder (ANSD). Background: ECochG is an emerging technique for predicting outcomes in CI recipients. Its utility may be increased by separating the cochlear microphonic (CM), produced by hair cells, from the ANN, the evoked potential correlate of neural phase-locking, which are mixed in the ongoing portion of the response to low frequency tone bursts. Methods: Responses to tone bursts of different frequency and intensities were recorded from the round window of pediatric CI recipients. Separation of the CM and ANN was performed using a model of the underlying processes that lead to the shapes of the observed waveforms. Results: Preoperative mean pure tone amplitudes of the included ANSD (n = 36) and non-ANSD subjects (n = 123), were similar (89.5 and 93.5, p = 0.1). Total of 1, 024 ECochG responses to frequency and intensity series were recorded. The mean correlation ( r ) between the input and the modeled signals was 0.973 ± 0.056 (standard deviation). The ANN magnitudes were higher in the ANSD group (ANOVAs, F = 26.5 for frequency and 21.9 for intensity, df's = 1, p 's < 0.001). However, its relative contribution to the overall signal was lower (ANOVAs, F = 25.8 and 12.1, df = 1, p 's < 0.001). Conclusions: ANN was detected in low frequency ECochG responses but not high frequency responses in both ANSD and non-ANSD subjects. ANSD subjects, evidence of neural contribution in responses to low frequency stimuli was highly variable and often comparable to signals recorded in non-ANSD subjects. The computational model revealed that on average the ANN comprised a lower proportion of the overall signal than in non-ANSD subjects.