Ke Lin1, Stephanie A Lee, Walter E Zink. 1. Department of Radiology, New York University Langone Medical Center, NY, USA. kelinmd@gmail.com
Abstract
PURPOSE: Infarct volume ≥100 mL on diffusion weighted imaging (DWI) predicts symptomatic hemorrhagic transformation and poor outcome. Our aim was to determine the correlation between the Alberta Stroke Program Early CT Score (ASPECTS) and infarct volume and to identify the optimal value for describing infarcts ≥100 mL. METHODS: This was a retrospective study of acute infarcts isolated to the middle cerebral artery territory imaged by DWI <48 hours from ictus. Two neuroradiologists blinded to volumetric measurements assigned ASPECTS while a third observer used a semi-automated thresholding technique to determine infarct volume. Correlation of ASPECTS and infarct volume was determined using Spearman's rank coefficient (ρ). Receiver-operating characteristics (ROC) curve analysis was performed to identify the optimal ASPECTS for ≥100 mL. RESULTS: One hundred and fifty patients were evaluated; the median and range for infarct volumes were 32.3 and 10.0-277 mL, respectively. The median and range for ASPECTS were 7 and 1-9, respectively. A strong correlation was found with ρ=-.807 (P < .0001). 22 (14.7%) infarcts were ≥100 mL and the area under the ROC curve was .976 (P < .0001). The optimal ASPECTS was ≤3 with sensitivity and specificity of 77.3% and 97.7%, respectively. CONCLUSION: ASPECTS may serve as a surrogate marker of infarct extent on DWI.
PURPOSE:Infarct volume ≥100 mL on diffusion weighted imaging (DWI) predicts symptomatic hemorrhagic transformation and poor outcome. Our aim was to determine the correlation between the Alberta Stroke Program Early CT Score (ASPECTS) and infarct volume and to identify the optimal value for describing infarcts ≥100 mL. METHODS: This was a retrospective study of acute infarcts isolated to the middle cerebral artery territory imaged by DWI <48 hours from ictus. Two neuroradiologists blinded to volumetric measurements assigned ASPECTS while a third observer used a semi-automated thresholding technique to determine infarct volume. Correlation of ASPECTS and infarct volume was determined using Spearman's rank coefficient (ρ). Receiver-operating characteristics (ROC) curve analysis was performed to identify the optimal ASPECTS for ≥100 mL. RESULTS: One hundred and fifty patients were evaluated; the median and range for infarct volumes were 32.3 and 10.0-277 mL, respectively. The median and range for ASPECTS were 7 and 1-9, respectively. A strong correlation was found with ρ=-.807 (P < .0001). 22 (14.7%) infarcts were ≥100 mL and the area under the ROC curve was .976 (P < .0001). The optimal ASPECTS was ≤3 with sensitivity and specificity of 77.3% and 97.7%, respectively. CONCLUSION: ASPECTS may serve as a surrogate marker of infarct extent on DWI.
Authors: Julian Schröder; Bastian Cheng; Martin Ebinger; Martin Köhrmann; Ona Wu; Dong-Wha Kang; David S Liebeskind; Thomas Tourdias; Oliver C Singer; Soren Christensen; Bruce Campbell; Marie Luby; Steven Warach; Jens Fiehler; Jochen B Fiebach; Christian Gerloff; Götz Thomalla Journal: Stroke Date: 2014-10-14 Impact factor: 7.914