MAGNETIC-RESONANCE-IMAGING OF EPIDURAL AN D SUBDURAL SPINAL HEMATOMAS

Epidural and subdural spinal hematomas were previously diagnosed by my elography and computed tomography (CT). Recent reports indicate that n oninvasive detection is possible with magnetic resonance imaging. We r eport on nine patients who were investigated by magnetic resonance ima ging (MRI) prior to surgery for epidural and subdural spinal hematoma. The MR examinations were performed on 1.5-T and 1-T units. We used su rface coils and employed T1-, PD- and T2-weighted spin echo sequences and a T2-weighted gradient echo sequence. CT was available in four pa tients and myelography in two patients. Surgical correlation was avail able in all patients. The hematomas were located in the cervical spine (n = 2), thoracic spine (n = 6) and lumbar spine (n = 2). They were e pidural in five patients and subdural in four. Blinded reading correct ly identified all five epidural hematomas and three of the subdural he matomas; one subdural hematoma was misjudged as epidural. Peracute hem atomas (< 24 h) in three patients appeared isointense or slightly hype rintense on T1-weighted images and had mixed signal intensity on T2- a nd T2-weighted images. Acute hematomas (1-3 days) in four patients we re also isointense on T1-weighted images but were more hypointense on T2- and T2-weighted images. Chronic hematomas in two patients (7 days and 14 days) were hyperintense on all sequences. Differentiation betw een epi- and subdural hematomas required transverse T2-weighted gradi ent echo sequences. Our results underline that MRI at 1 and 1.5 T is c apable of identifying epidural and subdural spinal hematoma in the acu te and peracute stage. MRI is superior to CT and myelography for the d elineation of the craniocaudal extension in epidural and subdural spin al hematomas and should be the primary preoperative diagnostic method.