Article abstract-Data from 694 patients hospitalized with stroke were entered in a prospective, computer-based registry. Three hundred and sixty-four patients (53 percent) were diagnosed as having thrombosis, 215 (31 percent) as having cerebral embolism, 70 (10 percent) as having intracerebral hematoma, and 45 (6 percent) a s having subarachnoid hemorrhage from aneurysm or arteriovenous malformations. The 364 patients diagnosed as having thrombosis were divided into 233 (34 percent of all 694 patients) whose thrombosis was thought to involve a large artery and 131 (19 percent) with lacunar infarction. Many of the findings in this study were comparable to those in previous registries based on postmortem data. New observations include the high incidence of lacunes and cerebral emboli, the absence of a n identifiable cardiac origin in 37 percent of all emboli, a nonsudden onset in 21 percent of emboli, and the occurrence of vomiting at onset in 51 percent and the absence of headache a t onset in 67 percent of hematomas.

The speech disturbance resulting from infarction limited to the Broca area has been delineated; it differs from the speech disorder called Broca aphasia, which results from damage extending far outside the Broca area. Nor does Broca area infarction cause Broca aphasia. The lesions in 20 cases observed since 1972 were documented by autopsy, computerized tomography, or arteriogram; the autopsy records from the Massachusetts General hospital for the past 20 years and the published cases since 1820 were also reviewed. The findings suggest that infarction affecting the Broca area and its immediate environs, even deep into the brain, causes a mutism that is replaced by rapidly improving dyspraxic and effortful articulation, but that no significant distrubance in language function persists. The more complex syndrome traditionally referred to as Broca aphasia, including Broca's original case, is characterized by protracted mutism, verbal stereotypes, and agrammatism. It is associated with a considerably larger infarct which encompasses the operculum, including the Broca area, insula, and adjacent cerebrum, in the territory supplied by the upper division of the left middle cerebral artery.

No abstract

To determine the prevalence of radiologically evident carotid stenosis in patients with transient cerebral ischemic attacks, we analyzed 95 consecutive hospitalized patients who during a two-year period had appropriate symptoms and also underwent angiography. Pure transient hemisphere symptoms affected 52 patients, pure monocular blindness occurred in 33, and 10 experienced each type of attack separately. Tight stenosis (less than or equal to 2 mm) or occlusion was present in 49 patients (52 per cent). Thirteen patients showed intracranial-branch occlusion, nine of whom had no notable stenosis. Only two clinical transient ischemic attack features correlated with angiographic findings: in transient hemisphere attacks lasting for one hour or longer, the carotid arteries revealed no notable stenosis (0.05 less than P less than 0.1); and separate hemisphere and ocular attacks in the same patient correlated with tight carotid stenosis. On the basis of the angiographic findings, the study indicates there are several distinct groups of patients with carotid transient ischemic attacks.