Abstract

P49 Background-Objectives: Cerebral ischemia results in heterogeneous tissue changes with distinct MRI features that differ from those of normal tissue. Segmentation of the ischemic lesion by multiparameter MRI generates different tissue clusters reflecting various degrees of tissue damage. Defining the MRI features of ischemic tissue that is capable of recovery could aid in the selection of patients for acute phase treatment. We performed this study in order to define the MRI characteristics of ischemic lesion clusters [tissue signatures (TS)] that have potential for recovery by using the computer segmentation algorithm Iterative Self-Organizing Data Analysis Technique (ISODATA). Methods: We included patients with acute ischemic stroke (IS) who received no thrombolysis and completed the multiparameter MRI studies at acute (≤12h) and chronic (≥3 months) time points. The MRI protocol included DWI, T2WI and T1WI. Post-processing analysis by ISODATA segmented the lesion into clusters and assigned a TS number to each based on the Euclidean distance between the the cluster and the normal tissue centers. White matter is represented by TS1,CSF by TS12, and all other clusters range between these two. We then identified acute time point TS that recovered (normal TS 1) by loading the chronic T2 lesion region of interest (ROI) onto the acute ISODATA image of the corresponding slice and location. ADC and T2 ratios were calculated by loading TS ROI onto ADC & T2 maps. Results: We studied 10 patients (age 56±14ys). We identified TS 2–8 in the acute ischemic lesions. TS 2 (n=1), TS 3 (n=1), TS 4 (n=9), TS 5 (n=4) and TS 6 (n=6) consisted part of the acute ischemic lesion that resolved by the chronic time point. TS ≥7 did not recover. The characteristics of the TS that recovered were: rADC = 0.738 (±0.137) and rT2 = 1.206 (±0.166). Conclusions: These preliminary data indicate that ISODATA-defined tissue signatures ≤6, consisting of low rADC and elevated rT2, still have the potential to recover. Consequently, ischemic lesions revealed by DWI and T2 are not necessarily irreversible. These findings should be considered in the design of MRI-based acute stroke treatment trials.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call