Clinical applicability of a preoperative angular insertion depth prediction method for cochlear implantation

Abstract

Objective: Evaluation of the accuracy and clinical applicability of a single measure cochlear implant angular insertion depth prediction method. Background: Cochlear implantation outcomes still vary extensively between patients. One of the possible reasons could be variability in intracochlear electrode array placement. For this reason, single measure methods were suggested to preoperatively predict angular insertion depths. Based on a previously performed accuracy study in human temporal bones, we were interested in determining the extent to which the method could be applied in a clinical setting. Methods: A retrospective analysis was performed on pre- and postoperative radiographic images of 50 cochlear implant recipients. Preoperatively predicted angular insertion depths were compared with angular insertion depths measured on postoperative ground truth. The theoretical prediction error was computed under the assumption that all achieved insertions were matching the preoperatively assumed linear insertion depth. More importantly, the clinical prediction error was assessed using two different software tools performed by three experienced surgeons. Results: Using the proposed method we found a theoretical prediction error of 5 degrees (SD¼41 degrees). The clinical prediction error including the cases with extracochlear electrodes was 70 degrees (SD¼96 degrees). Conclusions: The presented angular insertion depth prediction method is a first practical approach to support the preoperative selection of cochlear implant electrode arrays. However, the presented procedure is limited in that it is unable to predict the occurrence of insertion results with extracochlear electrodes and requires user training.