Abstract TP62: Acceleration-Selective Asl Perfusion Imaging as an Alternative for Dsc in an Acute Stroke Population

Abstract

Introduction: Differentiation of core, penumbra, and benign oligemia (based on capillary perfusion) is of increasing importance in treating acute ischemic stroke (AIS). Both CT and DSC perfusion MRI rely on similar kinetics which may be influenced by arrival delay. Arterial spin labeling (ASL) MRI serves as a non-invasive alternative but similarly suffers from delayed arrival. Here we explore acceleration-selective ASL (accASL), a method not dependent on arrival delay, for identifying regions of critically low perfusion. Methods: Forty-four patients with AIS and pre-treatment MRI including DWI, DSC and accASL MRI were included in this study. Perfusion and diffusion lesions were qualitatively and quantitatively (i.e. through manual delineation) rated on blinded DSC-DWI and AccASL-DWI image pairs. The mean transit time (MTT) map was used for DSC perfusion evaluation. Results: Only 43 patients were included in our analysis due to uninterpretable MTT maps in one patient. A perfusion lesion was identified on 38 of the MTT maps and 32 of the accASL images. The mean perfusion lesion volume was larger (p<0.001) on MTT than on the accASL images; 120 (±103) versus 32 (±52) ml, respectively. The mean lesion on DWI was 10 (±13) ml, and accASL detected perfusion deficit related better to DWI lesion than MTT (R 2 of 0.20 and 0.08, respectively). MTT identified a perfusion lesion in 8 patients where accASL did not. The median MTT perfusion lesion in these patients was 103 (3-198) ml and the median DWI lesion volume was 2 (0-7) ml. Two patients demonstrated a perfusion deficit on the accASL images (5 and 49 ml) but not on the MTT images, the DWI lesion volume in these patients was 1 and 2 ml. Conclusion: AccASL could identify a perfusion deficit in more than 75% of patients with a perfusion deficit on the MTT maps. Lesions on accASL images were closer in volume to those seen on DWI, suggesting a more specific measure of ischemia that is less influenced by delays due to proximal occlusion or stenosis. Future work will investigate the relation between accASL perfusion deficit and final infarct.