Accuracy of computed tomography detection of superior canal dehiscence

High-resolution temporal bone computed tomography (CT) may erroneously demonstrate a superior semicircular canal dehiscence (SSCD) where none exists and inaccurately display the size of a dehiscence. CT is an integral component of the diagnosis of SSCD. The prevalence of dehiscence as measured on computed tomographic scan is approximately eightfold higher than that on histologic studies, suggesting that CT may have a relatively low specificity for identifying canal dehiscence. This, in turn, can lead to an inappropriate diagnosis and treatment plan. We quantified the accuracy of CT in identifying a dehiscence of the superior semicircular canal in a cadaver model using microCT as a gold standard. The superior canals of 11 cadaver heads were blue lined. Twelve of the 22 ears were further drilled to create fenestrations of varying sizes. Heads were imaged using medical CT, followed by microCT scans of the temporal bones at 18-µm resolution. Diagnosis of dehiscence and measurements of dehiscence size were performed on clinical CT and compared with that of microCT. Clinical CT identified 7 of 8 intact canals as dehiscent and tended to overestimate the size of smaller fenestrations, particularly those surrounded by thin bone. These findings confirm that medical CT cannot be used as the exclusive gold standard for SSCD and that, particularly for small dehiscences on CT, clinical symptoms must be clearly indicative of a dehiscence before surgical treatment is undertaken. Preoperative counseling for small dehiscences may need to include the possibility that no dehiscence may be found despite radiologic evidence for it.