Comparison of computer tomography and magnetic resonance tomography in thediagnosis of intracerebral hemorrhage

Background and Purpose: Stroke symptoms are caused in 10 to 15% by intracer ebral hemorrhage. From the clinical examination it is often impossible to d ifferentiate intracerebral hemorrhage from cerebral ischemia. To exclude in tracerebral hemorrhage as the cause of clinical symptoms a CT is usually pe rformed. The aim of our study was a direct comparison of the sensitivity of Computed Tomography and MRI using different MR sequences for the detection of acute intracerebral hemorrhage. Methods: In 8 male Wistar rats intracerebral hemorrhage was induced by infu sion of collagenase into the caudate nucleus. After 1 hour the brains were subsequently imaged with CT and MRI using T2- and T1-weighted Spin Echo seq uences, diffusion-weighted sequences,T2*-weighted gradient echo sequences a nd FLAIR-sequences. Visibility of the intracerebral hemorrhage was examined using a scoring system from 1=not visible to 5=excellent visible. Finally, the intracerebral hemorrhage was verified by histological staining. Results: In all animals, intracerebral hemorrhage was visible in T2*-weight ed gradient echo and diffusion weighted MR images 1 h after infusion of col lagenase. T2- and PD-weighted SE images were positive in 7/8 rats. T1-weigh ted images revealed signal changes in 5/8 rats, and FLAIR sequence was posi tive in 8/8 rats. In CT intracerebral hemorrhage was only visible in 3/8 ra ts. When measuring the increase of Hounsfield units within the suspected he misphere we saw a mean increase of 7% compared to the normal hemisphere in 3/8 rats. Conclusions: In this animal model,T2*-weighted magnetic resonance imaging p roved to be the most sensitive imaging modality in the detection of acute i ntracerebral hemorrhage and is by far more sensitive than CT.