[CT differential diagnosis and findings in intracerebral haemorrhage (author's transl)]

CT in the diagnosis of cerebral vascular malformations

In 72 patients with spontaneous intracranial haemorrhage and vascular malformation, 27 of 49 arterial aneurysms were diagnosed by CT (the smallest one of 4mm diameter being stalked), as were all (18) of the arteriovenous aneurysms (angiomas), but of the venous malformations (5) only 2 (aneurysms of the vein of Galen) were so diagnosed. In the 25 patients with spontaneous subarachnoid haemorrhage, in whom vascular malformation had not been diagnosed through angiography or CT, CT showed the position and extent of the parenchymatous lesion or the existence of blood in the subarachnoid space or ventricules. Finally, in 15 patients with subjective or neuropsychiatric disturbances, 9 arterial and 6 arteriovenous aneurysms were diagnosed by CT and were verified by angiography, which would probably not have been performed if CT had not been performed. Thus it is clear that vascular malformations are often diagnosed by CT. In many cases information is revealed which would not be suspected with angiography, while in other cases angiography is more selective and accurate.

Follow-up of spontaneous intracerebral haemorrhage by computed tomography

The follow-up of 89 patients with spontaneous intracerebral haematomas was studied by computed tomography (CT). The aetiology of the haemorrhage was found to be hypertension in 52 cases, vascular malformation in 14 cases, and two haematomas were due to anticoagulant therapy. The lesion visible on CT consisted of an area of increased density surrounded by a rim of low attenuation. Most cases showed a progressive decrease in size of the total lesion. Exceptions to this pattern were observed, which may lead to the misdiagnosis of a tumour. The central area of increased density decreased in size much more quickly than the total lesion. Signs of a mass effect were observed up to the ninth week in the larger lesions. The resorption of blood, measured by the decrease in diameter of the central high attenuation area, was not dependent on the site or original size of the haematoma, but was slower in hypertensive haemorrhage than in haemorrhage due to vascular malformation. Contrast enhancement was observed up to the ninth week and was ring-like in most cases. Our data confirm the well-known finding that the bleeding cannot be demonstrated by examination of the CSF if the blood has not reached the ventricular system or the subarachnoid space.