To define a minimal set of descriptive parameters for cochlear morphology and study its influence on the cochlear implant electrode position in relation to surgical insertion distance.
Cochlear morphology and electrode position were analyzed using multiplanar reconstructions of the pre- and postoperative CT scans in a population of 336 patients (including 26 bilaterally implanted ones) with a CII HiFocus1 or HiRes90K HiFocus1J implant. Variations in cochlear diameter and cochlear canal size were analyzed. The relationship between the outer and inner walls was investigated. Size differences based on sex, age, and ear side were investigated using linear mixed models. Two new methods, spiral fitting and principal component analysis, were proposed to describe cochlear shape, and the goodness of fit was investigated. The relationship between cochlear shape and electrode position, in terms of modiolus proximity and insertion depth, was analyzed using clustering, one-way analysis of variance (ANOVA) and simple linear regression analysis.
Large variations in cochlear morphology were found, with cochlear canal sizes ranging from 0.98 to 2.96 mm and average cochlear diameters between 8.85 and 5.92 mm (with standard deviations of around 0.4 mm). The outer and inner walls were significantly correlated (p < 0.01), and a size difference of 4% in favor of males was found. Spiral fitting shows good alignment of the true measurements, with residuals having a mean of 0.01 mm and a standard deviation of 0.29 mm. Principal component analysis (PCA) showed that the use of one component, which describes size, is sufficient to explain 93.6% of the cochlear shape variance. A significant sex difference was also found with spiral fitting and PCA. Cochlear size was found to have a significant influence on modiolus proximity and insertion depth of the electrode (p < 0.01). Cochlear size explained around 13% of the variance in electrode position. When cochlear size was combined with surgical insertion, more than 81% of the variance in insertion depth can be explained.
This study demonstrates a large variety in cochlear morphology, which significantly impacts electrode position in terms of modiolus proximity and insertion depth. The effect size is, however, relatively small compared with surgical insertion distance. PCA is shown to be an accurate reduction method for describing cochlear shape.