Diagnostic maxillofacial coronal images reformatted from helically acquired thin-section axial CT data.

Abstract

This study was undertaken to determine the sensitivity and specificity of coronal images reformatted from helical thin-section axial CT data obtained for the evaluation of maxillofacial fractures. Multiple fractures were created in nine cadaver heads by blunt trauma and were then evaluated using a late-generation helical CT scanner. Two neuroradiologists then independently evaluated the axial and reformatted coronal maxillofacial images. Subsequently, they reviewed the axial and direct coronal CT images, which were considered the criterion standard. A total of 87 fractures were identified. An experienced neuroradiologist failed to identify one displaced fracture and two nondisplaced fractures when evaluating the reformatted coronal and direct axial images for an overall sensitivity of 97%. A less experienced neuroradiologist failed to identify a total of five minimally displaced or nondisplaced fractures for an overall sensitivity of 94%. For each radiologist, no significant difference in the time required to interpret the direct versus the reformatted coronal images was seen. Interpretation of axial and reformatted coronal images resulted in accurate identification of displaced maxillofacial fractures in cadavers. This study suggests that the added cost and radiation exposure associated with incremental direct coronal CT may not be necessary for detection of clinically significant maxillofacial fractures and that further evaluation of this protocol in live trauma patients is warranted. However, because nondisplaced fractures were not routinely detected using reformatted coronal images, physical examination and clinical suspicion will still also remain necessary to determine the need for further imaging.