Multivariate analysis of the exact effects of scalar position and insertion angle on speech discrimination

Abstract

Objective: Several studies examined the influence of cochlear morphology on scalar position of the electrode array and rate of dislocation. Furthermore, researchers described better speech discrimination for patients with electrode arrays positioned in scala tympani but in small study cohorts. The aim of this study is to examine the exact impact of scalar position, dislocation and angular insertion depth on postoperative speech perception.

Design: We identified the patients (n = 531) implanted between 2003 and 2018 with a Contour Advance electrode array (Cochlear^TM) inserted via cochleostomy by a retrospective review of the Cochlear Implant Database and analyzed the postoperative imaging by cone beam computed tomography and the audiological protocol via a multivariate nonparametric analysis.

Results: The multivariate nonparametric analysis of this study shows, that the dislocation of the electrode array and the insertion angle leads to no significant different postoperative speech discrimination results. Nevertheless, we could calculate a statistically significant amount of reduced speech recognition for monosyllables for primary scala tympani vs. scala vestibuli insertions (7.6%).

Conclusion: This study, based on one of the largest study cohorts published to date, demonstrates reduced speech recognition for scala vestibuli insertions compared to scala tympani insertions. Insertion into the scala vestibuli results in a 7.6% decrease in speech discrimination for monosyllables.

Keywords: cochlear coverage; cochlear implant; perimodiolar array; scala tympani; scala vestibuli; scalar position; speech perception.

Figure 1

Distribution of cochlear diameter A measured as the largest distance from the round window through the modiolus, and the perpendicular diameter B as established by Escudé et al. (7).

Figure 2

Distribution of the scalar position for the included Contour Advance (CA) electrode arrays (n = 531), (Cochlear^TM). (T = scala tympani, TD = dislocated out of scala tympani, V = scala vestibuli, VD = dislocated out of scala vestibuli).

Figure 3

Distribution of cochlear coverage (in°) for the included electrode array (n = 531).

Figure 4

(a) Presentation of the mixed effect model with presentation level as a known factor and modeling the asymptotic value (Asym) for speech discrimination in monosyllables. While we could not find significant influences for age (a), dislocation and coverage, primary scalar insertion (T vs. V) demonstrates significant influence (b). *p < 0.05.