Introduction: Cranial CT has been the most extended evaluation means for patients suffering head trauma. However, it has low sensitivity in the identification of diffuse axonal injury and posterior fossa lesions. Cranial MR is a potentially more sensitive test but difficult to perform in these patients, a fact that has hampered its generalised use.
Objective: To compare the identification capability of traumatic intracranial lesions by both diagnostic tests in patients with moderate and severe head injury and to determine which radiological characteristics are associated with the presence of diffuse injury in MR and their clinical severity.
Material and methods: 100 patients suffering moderate or severe head injury to whom a MR had been performed in the first 30 days after trauma were included. All clinical variables related to prognosis were registered, as well as the data from the initial CT following Marshall et al., classification. The MR was blindly evaluated by two neuroradiologists that were not aware of the initial CT results or the clinical situation of the patient. All lesions were registered as well as the classification following the classification of lesions related to DAI described by Adams et al. CT and MR findings were compared evaluating the sensitivities of each test. Factors related to the presence of diffuse injury in MR were studied by univariate analysis using chi2 test and simple correlations.
Results: MR is more sensitive than CT for lesions in cerebral white matter, corpus callosum and brainstem. It also detects a greater number of cerebral contussions. The presence of diffuse axonal injury depends on the mechanism of the trauma, being more frequent in higher energy trauma, specially in traffic accidents. Among the radiological characteristics associated to DAI the most clearly related is intraventricular haemorrhage. The presence of a deeper injury and a higher score in the scales of Adams is associated with a lower score in the GCS and motor GCS, and so with a worse level of consciousness and bigger severity of injury, confirming Ommaya's model.