Cochlear Implants (CIs) are a successful management option for individuals with severe-to-profound sensorineural hearing loss who cannot benefit from hearing aids.1,2 Technological advancements have greatly expanded CI candidacy to include other patients, such as adults with eligible degrees of hearing loss and children with congenital anomalies of the inner ear and auditory nerves.3–6 However, many factors affect the postoperative auditory performance of CI recipients, including hearing status before implantation, age at implantation, association with any other disabilities, type and length of the electrode used, and position of placement inside the cochlea.7,8 Additionally, human cochleae exhibit inter-individual variations in morphology, coiling characteristics, and size.9–11 These anatomical variations are associated with considerable differences in the tonotopic organization of the organ of corti along the cochlear duct. Hence, comprehensive preoperative planning for CI, primarily relying on radiological evaluation of the brain and temporal bone, is imperative. This evaluation helps surgeons assess the cochlear anatomy, identify any anatomical abnormalities in the ear, anticipate any potential factors that may contribute to surgical complications, and predict the type and optimal placement of the electrode array.12,13

Preoperative imaging is fundamental in determining CI candidacy. It is a common practice to assess the cochlear anatomy, identify congenital anomalies, plan surgery, customize the intraoperative surgical approach, and ensure proper electrode placement.14 However, no consensus currently exists regarding the preferred preoperative radiological evaluation modality. Both High-Resolution Computed Tomography (HRCT) and Magnetic Resonance Imaging (MRI) can be used, individually or in combination, depending on institutional guidelines and surgeon preferences.15 HRCT provides detailed visualization of cochlear bony structures, including anomalies of the bony labyrinth and cochlear sclerosis,16 but may miss up to 10% of soft tissue labyrinth abnormalities such as fibrous obliteration of the cochlea, membranous labyrinth, and endolymphatic sac abnormalities.17 In contrast, MRI allows the visualization of the fluid content of the membranous labyrinth, vestibulocochlear nerve hypoplasia or aplasia, cerebellopontine angle, cochlear fibrosis, and inflammation.14,18 MRI does not expose patients to radiation, addressing a notable concern associated with HRCT, particularly pediatric patients. CT and MRI can be considered complementary radiological modalities in preoperative CI planning.

OTOPLAN® is an otological surgical planning software designed to integrate data from both CT and MRI scans, providing surgeons with crucial information regarding various cochlear parameters.19–21 Initially, this software required manual labeling of critical cochlear structures (diameter, width, and height), followed by automated computation of Cochlear Duct Length (CDL).22 Previous studies have demonstrated the comparable performance of MRI and CT scans using the software's manual option.20,23 Nonetheless, one of the common challenges in manual radiological evaluation is the inconsistency in reporting radiological findings owing to inter-rater variability, which could be due to the levels of expertise, training backgrounds, and individual biases.20–22,24 Hence, a fully Automated algorithm (AUTO) was recently introduced into the software, which has been evaluated only on data integrated from CT scans.25

Although MRI can provide all the necessary data for CI planning in addition to being radiation-free compared to CT,26 no previous studies have assessed automated measurements obtained from MRI scans. Thus, this study aims to evaluate the precision of automated cochlear measurements obtained from MRI scans compared to those derived from manual workflows using MRI and CT scans.

A recent version of OTOPLAN® introduces automatic cochlear parameter measurements, which are compatible with both CT and MRI (AUTO).19 This study aimed to validate AUTO by comparing the fully automated cochlear measurements obtained from MRI scans (MRI-Auto) with those extracted from the manual option using CT and MRI scans. To our knowledge, this is the first study to validate MRI-Auto as a preoperative cochlear implant planning tool.

Radiological evaluation is crucial in preoperative preparation for CI. Although high-resolution CT is considered the gold standard for inner ear imaging, the radiation dose used in CT raises concerns, mainly when cumulative radiation exposure occurs at a young age.29,30 MRI in preoperative CI planning offers several advantages, including a comprehensive anatomical view with excellent inner ear soft tissue contrast without harmful ionizing radiation.31 Interestingly, comparable performances of MRI and CT in the preoperative evaluation of CI have been reported.13 Together, these findings underline the significance of validating AUTO using MRI as an alternative to CT.

We started this study by recording two different readings for each cochlear parameter using manual features on CT and MRI images of the same patient group and one reading from MRI-Auto. Our cochlear parameter measurements were as follows: the average A value (cochlear diameter) was 8.73 for MRI-Auto and 8.79 mm for both manual options using CT and MRI. These values fall within the range previously reported in a Review Article, which showed that the A values ranged from 7.5 to 10.5 mm, with an average value of 9.0 mm.28 Our findings demonstrated that the average B value (cochlear width) for MRI-Auto and manual CT and MR measurements were 6.62 mm and 6.65 mm, respectively. The average CDL calculated by MRI-Auto was 34.56 mm, and those for the manual options using CT and MRI were 34.77 and 34.94, respectively. Our B values and CDL measurements were comparable to those of prior studies, one of which used CT scans to validate AUTO.25,32,33 It should be noted that these studies performed cochlear measurements using older versions of the software that only have manual features.

Our comparative analysis demonstrated no significant differences in the measurements of A values, B values, and CDL using the three measurement approaches. More specifically, MRI-Auto assessment was comparable to manual CT and MRI assessment. These results are consistent with previous studies that considered the difference between CT and MRI cochlear metrics to be non-significant.20,33,34 However, one study reported significant differences in CDL measurement, with greater values detected by MRI than by flat-panel volume CT measurements.35 However, the difference was minor and of questionable clinical significance.

As previously noted in two different studies.20,25 Our results revealed excellent inter-observer reliability between the two readings of A values, B values, and CDL performed by two experienced physicians using manual CT (A value: ICC = 0.87, B value: ICC = 0.87, CDL: ICC = 0.92) and MRI (A value: ICC = 0.84, B value: ICC = 0.86, CDL: ICC = 0.90). The high reliability detected between cochlear metrics agrees with other studies, which reported even better inter-rater reliability when assessing the CDL.22,26

To gain further insight into the degree of precision and accuracy of the measurements obtained using MRI-Auto, the Concordance Correlation Coefficients (CCC) were calculated compared to measurements derived manually using MRI and CT scans. Our findings revealed that the concordance between the CT and MRI-Auto measurements was higher than between the CT and manual MRI measurements. These findings indicate that MRI-Auto outperforms the manual MRI workflow when evaluating cochlear parameters. The superior performance of MRI-Auto can be justified by its higher accuracy and precision, which offers finer measurements and a more accurate evaluation of cochlear parameters than a manual workflow. Moreover, automation reduces the possibility of subjective interpretation, which may arise from differences in the user experience. Additionally, the Auto feature can save time during the measurement steps and thus facilitate the planning process. Notably, the degree of concordance varied with different cochlear parameters, ranging from good to moderate and satisfactory degrees for A, B, and CDL, respectively.

This study has some limitations, such as the small sample size, and all included MRI scans of normal anatomical structures of the cochleae without any recorded malformations. Therefore, further investigations involving another cohort of patients with cochlear malformations are essential for a comprehensive assessment.

In conclusion, automatic assessment of cochlear parameters using MRI-Auto using OTOPLAN®was comparable to manual CT and MRI assessment. As no clinically significant difference was detected, using MRI-Auto in preoperative planning for CI appears appealing. Moreover, MRI-Auto showed superior performance in evaluating cochlear metrics compared to the manual MRI workflow. This MRI-Auto method for pre-operative planning could be an important step toward radiation-free cochlear implantation.

No funding from any organization was received for this work.