CT-Based Virtual Endoscopy to Predict Round Window Accessibility During Cochlear Implant

Objectives: To investigate the validity of the CT-based virtual endoscopy (VE) in predicting round window (RW) accessibility during cochlear implant (CI). Patients: A prospective case series study included 40 CI patients. Intervention: Radiological rating was performed by 2 CI surgeons using CT scan DICOM data and OsiriX software, through which virtual mastoidectomy and posterior tympanotomy were simulated to predict RW accessibility, which was graded as favorable or unfavorable. An intraoperative rating was performed, and RW visibility was classified as visible or nonvisible. Main outcome measures: We established a correlation between preoperative virtual RW accessibility and intraoperative RW visibility. Inter-rater and intrarater reliability were done for CT VE. Results: The mean age was 24 months, with males accounting for 58% of cases and right ears being implanted in 88% of cases. Intraoperative rating: we reported 29 cases with visible RW and 11 cases with nonvisible RW. Radiological rating: rater 1 documented 28 cases with favorable RW accessibility and 12 cases of unfavourability. Rater 2 documented 27 cases with favorable RW accessibility and 13 cases that were unfavorable. A correlation was established between both methods and revealed an average accuracy of 96%. Inter-rater reliability had an intraclass correlation coefficient of 0.971. Conclusion: We reported a novel method for predicting RW accessibility using 3D CT-based virtual endoscopy. This method is valid, reliable, reproducible, and feasible with a high accuracy rate.