Cochlear length determination using Cone Beam Computed Tomography in a clinical setting

Indications for cochlear implants are determined by audiological and medical considerations. Clinical imaging is therefore an integral element for anatomical evaluation in terms of medical considerations. Several authors have discussed the variability of cochlear shape, especially cochlear length. Cochlear length is, however, an increasingly recognized parameter in terms of preoperative evaluation. This study introduces a methodology to determine individual cochlear length in clinical setting by using Cone Beam Computed Tomography. Cochlear length determination was performed retrospectively with an OsiriX curved 3D Multiplanar Reconstruction tool on subjects who underwent temporal bone imaging from January 2011 to February 2013. Cochlear length was defined as the spiral route from the center-distal point of the bony round window along the lateral wall towards the helicotrema, which is the endpoint of the measurement. Cochlear length was measured in 436 temporal bones (218 left ears, 218 right ears, 218 subjects). The mean cochlear length was 37.6 mm (SD: +/- 1.93 mm), median was 37.6 mm, range 32-43.5 mm. The cochlear length had a normal distribution. A significant difference was found between cochlear length by gender (p < .0001), but not between the left and right cochlea (p = .301) or according to age. Consideration of the cochlear length in clinical data may be an insufficiently represented parameter in cochlear implant treatment. Literature shows the impact of electrode insertion depth on residual hearing preservation and speech performance. Individual evaluation of the cochlear implant electrode choice may be the next step in personalized cochlear implant treatment as a valuable addition to existing audiological and surgical evaluation. The cochlear length determination methodology presented herein is a reproducible and clinically available parameter. Indeed, revealing a significant cochlear length span width, especially according to gender differences, may be assumed as hardly ignorable. (C) 2014 Elsevier B.V. All rights reserved.