Precise Evaluation of the Cochlear Duct Length by Flat-panel Volume Computed Tomography (fpVCT)-Implication of Secondary Reconstructions.

HYPOTHESIS: Flat-panel volume computed tomography (fpVCT) and secondary reconstruction allow for more accurate measurements of two-turn length (2TL), cochlear duct length (CDL), and angular length (AL). BACKGROUND: Cochlear geometry is a controversially debated topic. In the meantime, there are many different studies partly reporting highly divergent values. Our aim is to discuss the differences and to propose a radiological possibility to improve cochlear measurements using 3D-curved multiplanar reconstruction and fpVCT. METHODS: Performing different image modalities and settings, we tried to find a clinically usable option that allows for a high degree of accuracy. Therefore, we tested them against reference values of high-definition micro-computed tomography. RESULTS: Comparison of 99 µm slice thickness secondary reconstruction of fpVCT and reference showed no significant differences for 2TL and CDL (p ≥ 0.05). Accordingly, ICC (intraclass correlation) values were excellent (ICC ≥ 0.75; lower limit of confidence interval [CI] ≥ 0.75; Cronbach's alpha [α] ≥ 0.9). Evaluating AL, there was a significant difference (difference: -17.27°; p = 0.002). The lower limit of the CI of the ICC was unacceptable (ICC = 0.944; lower limit of CI = 0.248; α = 0.990). Regarding the Bland-Altman plots, there were no clinically unacceptable errors, but a systematic underestimation of AL. CONCLUSION: Secondary reconstruction is a suitable tool for producing reliable data that allow the accurate measurement of 2TL and CDL. The option of generating these reconstructions from raw data limits the need for higher radiation doses. Nevertheless, there is an underestimation of AL using secondary reconstructions.