Hypothesis: Automated processing of postoperative clinical cone-beam CT (CBCT) of cochlear implant (CI) patients can be used to accurately determine electrode contacts and integrated with an atlas-based mapping of cochlear microstructures to calculate modiolar distance, angular insertion distance, and scalar location of electrode contacts.
Background: Hearing outcomes after CI surgery are dependent on electrode placement. CBCT is increasingly used for in-office temporal bone imaging and might be routinely used for pre- and post-surgical evaluation.
Methods: Thirty-six matched pairs of pre- and postimplant CBCT scans were obtained. These were registered with an atlas to model cochlear microstructures in each dataset. Electrode contact center points were automatically determined using thresholding and electrode insertion parameters were calculated. Automated localization and calculation were compared with manual segmentation of contact center points as well as manufacturer specifications.
Results: Automated electrode contact detection aligned with manufacturer specifications of spacing and our algorithms worked for both distantly- and closely spaced arrays. The average difference between the manual and the automated selection was 0.15 mm, corresponding to a 1.875 voxel difference in each plane at the scan resolution. For each case, we determined modiolar distance, angular insertion depth, and scalar location. These calculations also resulted in similar insertion values using manual and automated contact points as well as aligning with electrode properties.
Conclusion: Automated processing of implanted high-resolution CBCT images can provide the clinician with key information on electrode placement. This is one step toward routine use of clinical CBCT after CI surgery to inform and guide postoperative treatment.
Objectives: Virtual reality (VR) simulation for patient-specific pre-surgical planning and rehearsal requires accurate segmentation of key surgical landmark structures such as the facial nerve, ossicles, and cochlea. The aim of this study was to explore different approaches to segmentation of temporal bone surgical anatomy for patient-specific VR simulation.
Methods: De-identified, clinical computed tomography imaging of 9 pediatric patients aged 3 months to 12 years were obtained retrospectively. The patients represented normal anatomy and key structures were manually segmented using open source software. The OTOPLAN (CAScination AG, Bern, Switzerland) otological planning software was used for guided segmentation. An atlas-based algorithm was used for computerized, automated segmentation. Experience with the different approaches as well as time and resulting models were compared.
Results: Manual segmentation was time consuming but also the most flexible. The OTOPLAN software is not designed specifically for our purpose and therefore the number of structures that can be segmented is limited, there was some user-to-user variation as well as volume differences compared with manual segmentation. The atlas-based automated segmentation potentially allows a full range of structures to be segmented and produces segmentations comparable to those of manual segmentation with a processing time that is acceptable because of the minimal user interaction.
Conclusion: Segmentation is fundamental for patient-specific VR simulation for pre-surgical planning and rehearsal in temporal bone surgery. The automated segmentation algorithm currently offers the most flexible and feasible approach and should be implemented. Further research is needed in relation to cases of abnormal anatomy.
Objectives: Patient-specific surgical simulation allows presurgical planning through three-dimensional (3D) visualization and virtual rehearsal. Virtual reality simulation for otologic surgery can be based on high-resolution cone-beam computed tomography (CBCT). This study aimed to evaluate clinicians’ experience with patient-specific simulation of mastoid surgery.
Methods: Prospective, multi-institutional study. Preoperative temporal bone CBCT scans of patients undergoing cochlear implantation (CI) were retrospectively obtained. Automated processing and segmentation routines were used. Otologic surgeons performed a complete mastoidectomy with facial recess approach on the patient-specific virtual cases in the institution’s temporal bone simulator. Participants completed surveys regarding the perceived accuracy and utility of the simulation.
Results: Twenty-two clinical CBCTs were obtained. Four attending otologic surgeons and 5 otolaryngology trainees enrolled in the study. The mean number of simulations completed by each participant was 16.5 (range 3-22). “Overall experience” and “usefulness for presurgical planning” were rated as “good,” “very good,” or “excellent” in 84.6% and 71.6% of the simulations, respectively. In 10.7% of simulations, the surgeon reported to have gained a significantly greater understanding of the patient’s anatomy compared to standard imaging. Participants were able to better appreciate subtle anatomic findings after using the simulator for 60.4% of cases. Variable CBCT acquisition quality was the most reported limitation.
Conclusion: Patient-specific simulation using preoperative CBCT is feasible and may provide valuable insights prior to otologic surgery. Establishing a CBCT acquisition protocol that allows for consistent segmentation will be essential for reliable surgical simulation.
Purpose: To develop an automated segmentation approach for cochlear microstructures [scala tympani (ST), scala vestibuli (SV), modiolus (Mod), mid-modiolus (Mid-Mod), and round window membrane (RW)] in clinical cone beam computed tomography (CBCT) images of the temporal bone for use in surgical simulation software and for preoperative surgical evaluation.
Methods: This approach was developed using the publicly available OpenEar (OE) Library that includes temporal bone specimens with spatially registered CBCT and 3D micro-slicing images. Five of these datasets were spatially aligned to our internal OSU atlas. An atlas of cochlear microstructures was created from one of the OE datasets. An affine registration of this atlas to the remaining OE CBCT images was used for automatically segmenting the cochlear microstructures. Quantitative metrics and visual review were used for validating the automatic segmentations.
Results: The average DICE metrics were 0.77 and 0.74 for the ST and SV, respectively. The average Hausdorff distance (AVG HD) was 0.11 mm and 0.12 mm for both scalae. The mean distance between the centroids for the round window was 0.32 mm, and the mean AVG HD was 0.09 mm. The mean distance and angular rotation between the mid-modiolar axes were 0.11 mm and 9.8 degrees, respectively. Visually, the segmented structures were accurate and similar to that manually traced by an expert observer.
Conclusions: An atlas-based approach using 3D micro-slicing data and affine spatial registration in the cochlear region was successful in segmenting cochlear microstructures of temporal bone anatomy for use in simulation software and potentially for pre-surgical planning and rehearsal.