Automated detection of electrically evoked stapedius reflexes (eSR) during cochlear implantation

Abstract

Abstract Introduction In cochlear implantation, objective fitting methods are needed to optimize audiological results in small children or patients with poor compliance. Intraoperatively measured electrically evoked stapedius reflexes (eSR) can be used as a marker for the patient’s discomfort level. The aim of this study was to develop and evaluate an automated detection method for eSR and to compare it to the detection rate of the surgeon and independent observers. Methods Cochlear implantation using a fully digital surgical microscope was performed. Movements of the stapedius tendon were recorded and analyzed by means of computer vision technique. Differences in eSR elicited by stimulating electrodes at different cochlear locations (basal, middle and apical) were analyzed. The eSR detection rate of the image processing algorithm was compared to the surgeon’s detection rate and to those of two less experienced observers. Results A total of 387 electrically impulses were applied. The stimulation of middle turn electrodes showed significantly higher detection rates (50.4%) compared to the basal (40.0%; p = 0.001) and apical (43.6%; p = 0.03) turn. The software identified significantly more of the applied stimuli (58.4%) compared to the surgeon (46.3%; p = 0.0007), the intermediate observer (37.7%; p < 0.0001) and the unexperienced observer (41.3%; p < 0.0001). Conclusion The feasibility of an automated intraoperative software-based detection of eSR is demonstrated. By improving the eSR detection methods and their clinical applicability, their utility in objective cochlear implant fitting may be substantially increased.