Page 403

Karger_ESC London_2013

London, United Kingdom 2013 Poster Session Red Cerebrovasc Dis 2013; 35 (suppl 3)1-854 403 223 Brain imaging Voxel based post-test probabilities maps to predict regions at risk of infarction following thrombolysis T.G. PHAN1, H. MA2, V. SRIKANTH3, B. CLISSOLD4, J. LY5 STROKE AND AGING RESEARCH, CLAYTON, AUSTRALIA1, STROKE AND AGING RE-SEARCH, CLAYTON, AUSTRALIA2, STROKE AND AGING RESEARCH, CLAYTON, AUS-TRALIA3, STROKE AND AGING RESEARCH, CLAYTON, AUSTRALIA4, STROKE AND AG-ING RESEARCH, CLAYTON, AUSTRALIA5 Background: The ischemic penumbra or salvageable tissue is defined by mismatch between the var-ious parameters of perfusion abnormalities on computer tomography (CT) perfusion. This volume based approach assumes that all regions have the same risk. However, regional variation in leptome-ningeal anastomoses may alter the regional risk of infarction. Hence a voxel based method may help to clarify the fate of tissue in different brain regions. We hypothesis that the likelihood of tissue in-farction differs depend on the brain region and illustrate this using a voxel based approach. Method: The inclusion criteria were: occlusion of middle cerebral artery (MCA) on CT angiography, CT perfusion (64 multirow detector/8 cm coverage) performed within 6 hours of stroke onset and patients receiving thrombolysis (n=30). Penumbra voxels were defined as voxels > 4 sec on mean transit time (MTT) map and <2.5 ml/100 g on cerebral blood volume (CBV) map. MCA infarct was defined on FLAIR images in patients with MCA occlusion. At a given voxel, we calculated the pos-itive likelihood ratio (PLR). Using a separate group of patients (n=28) with MCA occlusion, we cal-culate the pre-test odds of infarction. We combined the PLR map and the pre-test odds map to calcu-late the post-test probability of infarction following thrombolysis. Results: Perfusion abnormalities within the MCA territory were heterogenous, reflecting the differ-ent regions at higher risk of ischaemia. The post-test probability for the deep compartment (striato-capsular region and internal watershed region) of the MCA has probability > 0.80, the immediate region surrounding region has probability 0.40-0.79, follow by a region with probability between 0.1-0.39 while the more distal posterior region has probability <0.1. Discussion: The deep compartment of the MCA territory has probability for infarction despite re-ceiving thrombolysis. A voxel based approach may help to refine the zone from least likely to most likely to infarct. 224 Brain imaging Arterial spin labelling versus bolus-tracking CT and MR in hyper-acute ischemic stroke A. Bivard1, P. Stanwell2, N. Spratt3, C. Levi4, V. Krishnamurthy5, S. Davis6, M. Parsons7 University of Melbourne, Melbourne, AUSTRALIA1, University of Newcastle, Newcastle, AUS-TRALIA2, University of Newcastle, Newcastle, AUSTRALIA3, University of Newcastle, Newcas-tle, AUSTRALIA4, John Hunter Hospital, Newcastle, AUSTRALIA5, University of Melbourne, Melbourne, AUSTRALIA6, University of Newcastle, Newcastle, AUSTRALIA7 Arterial Spin Labeling (ASL) is a magnetic resonance perfusion imaging technique that does not require contrast administration and thus may be more practical in hyper-acute stroke than suscep-tibility- weighted bolus-tracking perfusion imaging (PWI). However, ASL requires validation and comparison to conventional bolus contrast perfusion imaging (MR and CT) in acute ischemic stroke. Methods: 55 patients with acute hemispheric ischemic stroke were imaged within 6 hours of symp-tom onset with perfusion CT followed by MRI including pASL, PWI and diffusion weighted imag-ing (DWI) .Receiver Operating Characteristic (ROC) Curve Analysis was used to test the accuracy of ASL in measuring the MRI DWI infarct core (b=1000), the PWI Tmax 6 seconds lesion, and the CTP Tmax 6 seconds lesion. Additionally, three neurologists were asked to independently rate the treatment suitability of patients based on randomly presented PWI, CTP and ASL maps and a Co-hen’s kappa coefficient was measured. Treatment criteria was set at an infarct core to penumbra ratio of >1:1.2 mismatch(MM) for all imaging modalities. Results: ASL measures of the acute perfusion lesion in combination with acute DWI overestimated ’tissue at risk’ when compared to PWI Tmax 6sec (AUC 0.72 SD 0.65 -0.81), mostly due to underestimation of blood flow in white matter. ASL had similar accuracy when compared to a CTP Tmax >6 sec threshold (AUC 0.74 SD 0.63-0.84). ASL lead to minimal change in automated MM classification based upon extent of PWI versus ASL defined tissue at risk and visual interpretation (Kappa=0.74). Discussion: pASL shows promise, however it overestimates hypoperfusion lesion volume due to low perfusion values in cerebral white matter. ASL is clinically useful, however the acquisition requires refining (e.g. with background sup-pression or longer ASL acquisitions)


Karger_ESC London_2013
To see the actual publication please follow the link above