Feasibility of three-dimensional ultrashort echo time magnetic resonance imaging at 1.5 T for the diagnosis of skull fractures

Abstract

To investigate the feasibility of ultrashort echo time (UTE) magnetic resonance imaging (MRI) for the diagnosis of skull fractures. The skull fracture models of ten Bama pigs and 364 patients with craniocerebral trauma were subjected to computed tomography (CT), UTE and conventional MRI sequences. The accuracy of UTE imaging in skull fracture diagnosis was analysed using receiver operating characteristic (ROC) curve analysis, McNemar's test and Kappa values. Differences among CT, UTE imaging and anatomical measurement (AM) values for linear fractures (LFs) and depressed fractures (DFs) were compared using one-way ANOVA and a paired-samples t-test. UTE imaging clearly demonstrated skull structures and fractures. The accuracy, validity and reliability of UTE MRI were excellent, with no significant differences between expert readings (P > 0.05; Kappa, 0.899). The values obtained for 42 LFs and 13 DFs in the ten specimens were not significantly different among CT, UTE MRI and AMs, while those obtained for 55 LFs and ten DFs in 44 patients were not significantly different between CT and UTE MRI (P > 0.05). UTE MRI sequences are feasible for the evaluation of skull structures and fractures, with no radiation exposure, particularly for paediatric and pregnant patients. Despite ionising radiation, CT is standard for skull fracture assessment. Conventional MRI cannot depict skull structures. 3D-UTE sequences clearly demonstrate skull structures and fractures. UTE plus conventional MRI are superior to CT in craniocerebral trauma assessment. Paediatric and pregnant patients will benefit from this imaging modality.