T2 imaging can assess ischemic insults as well as ADC imaging, but only 3.5 hours after the onset of ischemia. Assessment of edematous swelling (approximately 24.5% of total infarcted volume) demonstrates that ADC and therefore T2 imaging detect all the tissue that will become infarcted approximately 7 hours after occlusion. The spread of ADC and T2 abnormalities would therefore stop at approximately 7 hours, and any further increase in volume observed on the images would be mainly due to edematous swelling.

A new method for studying brain microcirculation is described. Both fluorescently labeled erythrocytes and plasma were visualized on-line through a closed cranial window in anesthetized rats, using laser-scanning two-dimension confocal microscopy. Video images of capillaries, arterioles, and venules were digitized off-line to measure microvessel diameter and labeled erythrocyte flow and velocity in parenchymal capillaries up to 200 microm beneath the brain surface. The method was used to analyze the rapid adaptation of microcirculation to a brief decrease in perfusion pressure. Twenty-second periods of forebrain ischemia were induced using the tour-vessel occlusion model in eight rats. EEG, arterial blood pressure, and body temperature were continuously controlled. In all conditions, labeled erythrocyte flow and velocity were both very heterogeneous in capillaries. During ischemia, capillary perfusion was close to 0, but a low blood flow persisted in arterioles and venules, while EEG was flattening. The arteriole and venule diameter did not significantly change. At the unclamping of carotid arteries, there was an instantaneous increase (by about 150%) of arteriole diameter. Capillary erythrocyte flow and velocity increased within 5 seconds, up to, respectively, 346 +/- 229% and 233 +/- 156% of their basal value. No capillary recruitment of erythrocytes was detected. All variables returned to their basal levels within less than 100 seconds after declamping. The data are discussed in terms of a possible involvement of shear stress in the reperfusion period.