We attempted to evaluate the diagnostic usefulness of the degree of perilesional edema around intracerebral hematoma in predicting the underlying cause.

This study included 54 patients with intracerebral hematoma for whom the underlying cause was confirmed by biopsy, radiological or clinical methods. Cases of subarachnoid hemorrhage, hemorrhagic transformation of cerebral infarction and intraventricular hemorrhage were excluded. The lesion size was defined as the average value of the longest axis and the axis perpendicular to this. The size of the perilesional edema was defined as the longest width of the edema. In all cases, the sizes of the lesion and edema were measured on the T2 weighted image. We defined the edema ratio as the edema size divided by the lesion size.

23 cases were diagnosed as intracerebral hemorrhage due to neoplastic conditions, such as metastasis (n=17), glioblastoma (n=5), hemangioblastoma(n=1). 31 cases were caused by non-neoplastic conditions, such as spontaneous hypertensive hemorrhage (n=23), arteriovenous malformation (n=4), cavernous angioma (n=3), and moya-moya disease (n=1). In fourteen cases, which were confirmed as malignant intracerebral hemorrhage, the edema ratio was more than 100%. Of the other cases, only 8 were confirmed as malignant intracerebral hemorrhage. It was found that the larger the edema ratio, the more malignant the intracerebral hemorrhage, and this result was statistically significant (p<0.001).

Measurement of perilesional edema and the intracerebral hematoma ratio may be useful in predicting the underlying causes.