The ASPECTS template is weighted in favor of the striatocapsular region Thanh G. Phan, a,b,e Geoffrey A. Donnan, a,e Masatoshi Koga, a,e L. Anne Mitchell, c,d Maurice Molan, c Gregory Fitt, c Winston Chong, f Michael Holt, f and David C. Reutens e, * a National Stroke Research Institute, Melbourne, Australia b Department of Medicine, Melbourne, Australia c Department of Radiology, Austin Health, Melbourne, Australia d University of Melbourne, Melbourne, Australia e Monash Institute of Neurological Disorders, Monash Medical Centre, Clayton Rd, Clayton, Victoria, Melbourne, Australia f Department of Radiology, Monash Medical Centre, Melbourne, Australia Received 16 October 2005; revised 22 December 2005; accepted 23 December 2005 Available online 28 February 2006 The extent of cerebral infarction correlates with increased risk of intracerebral hemorrhage (ICH) following recombinant tissue plas- minogen activator (rt-PA) administration. The Alberta Stroke Program Early CT Score (ASPECTS) is a widely used, validated method which assesses involvement of 10 selected regions of the MCA territory. An ASPECTS score >7 is associated with a higher risk of ICH following thrombolysis than lower scores. To understand the internal structure of the ASPECTS template better, we estimated the infarct volume corresponding to each region. We hypothesized that, in the ASPECTS scoring system, the striatocapsular region is weighted disproportion- ally. Four experienced radiologists rated individual ASPECTS regions on subacute CT images (day 5 – day 10) of 19 patients with MCA territory stroke. Infarct volume was determined from manual segmentation of infarcts on CT images. Linear regression was used to estimate the regional volume associated with each ASPECTS region. The ASPECTS regions are weighted unequally with the striatocapsular region accounting for 21% of the MCA territory infarct volume. Together, the 10 ASPECTS regions account for approximately 51% of the maximum MCA infarct territory volume. These findings should provide impetus for research to develop a scoring system explicitly based on regional hemorrhage risk as an aid to selecting patients for thrombolysis. D 2006 Elsevier Inc. All rights reserved. Keywords: Middle cerebral artery; Infarct; Regression; ASPECTS; Thrombolysis Introduction Stroke is the second commonest cause of disability in the world and a major cause of mortality (Bonita, 1992). Recent meta- analyses show that thrombolysis with recombinant tissue plasmin- ogen activator (rt-PA) reduces morbidity in selected patients with acute ischemic stroke (Hacke et al., 2004; Wardlaw et al., 2003). One concern with the use of this therapy is the risk of precipitating brain parenchymal hemorrhage, hence the interest in using imaging techniques to predict the likelihood of this complication. Throm- bolysis is then withheld from patients at high risk of hemorrhage if the imaging model can accurately predict this risk. Previous studies have proposed that the risk of intracerebral hemorrhage (ICH) following thrombolysis is increased in patients with middle cerebral artery (MCA) territory infarcts in whom visual inspection of acute CT scans indicates involvement of >1/3 of the arterial territory by infarction (Barber et al., 2000; Fiorelli et al., 1999; Larrue et al., 1997; Tanne et al., 2002). This prediction rule is currently used in MRI-based thrombolytic trials such as Desmoteplase In Acute ischaemic Stroke [DIAS] (Hacke et al., 2005). In this paper, we refer to this guideline as Fthe rule of 1/3_. Another bedside tool in prediction of hemorrhage, the Alberta Stroke Program Early CT Score (ASPECTS) system, was devised to help quantify early ischemic changes in the MCA territory on CT scans (Barber et al., 2000). The ASPECTS template comprises two slices, one at the level of the thalamus and striatum and one 2 cm superior to this level. One point is deducted for partial or total involvement by infarct in any of the 10 regions. According to its developers, an ASPECTS rating of 10 represents no visible infarction and a score of 0 represents diffuse ischemia throughout the MCA territory. ICH complicating thrombolysis is less likely to occur in patients with an ASPECTS rating between 8 and 10 in this patient group (Barber et al., 2000). However, the predictive value of a lower ASPECTS score for symptomatic thrombolysis-related 1053-8119/$ - see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2005.12.059 * Corresponding author. Fax: +613 9594 5662. E-mail address: david.reutens@med.monash.edu.au (D.C. Reutens). Available online on ScienceDirect (www.sciencedirect.com). www.elsevier.com/locate/ynimg NeuroImage 31 (2006) 477 – 481