hide
Free keywords:
magnetic resonance imaging; cerebral ischemia; stroke; diffusion; segmentation; reperfusion
Abstract:
Copyright 2002 Wiley-Liss, Inc.
Abstract:
The potential of multiparametric MRI parameters for differentiating between reversibly and irreversibly damaged brain tissue was investigated in an experimental model of focal brain ischemia in the rat. The middle cerebral artery (MCA) was occluded by intraluminal suture insertion for 60 or 90 min, followed by 4.5 h of reperfusion. The apparent diffusion coefficient (ADC) of brain water, T-1 and T-2 relaxation times, and CBFi, an MR-derived index of cerebral perfusion, were repeatedly measured and correlated with the outcome from the ischemic impact. A novel user-independent approach for segmentation of ADC maps into classes of increasing injury was introduced to define regions of interest (ROIs) in which these parameters were evaluated. MCA occlusion led to a graded decline of ADC, which corresponded with both the severity of flow reduction and an increase in T-1 and T-2 relaxation times. Removal of the suture led to a triphasic restitution of blood flow consisting of a fast initial rise, a secondary decline, and final normalization. Postischemic reperfusion led to a rise of ADC irrespective of the duration of ischemia. However, the quality of recovery declined with increasing severity of the ischemic impact. Throughout the observation time, T-1 and T-2 showed a continuous increase, the intensity of which correlated with the severity of ADC decline during ischemia. Particularly with longer ischemia time, elevated T-2 in combination with reduced ADC yielded a lower probability of recovery during recirculation, while intraischemic perfusion information contributed less to the prediction of outcome. In conclusion, the combination of MR parameters at the end of ischemia correlated with the probability of tissue recovery but did not permit reliable differentiation between reversibly and irreversibly damaged tissue.
