Abstract WMP116: Acute Infarct Volume in Childhood Stroke Can Be Accurately Estimated by Modelling Contraction of Chronic Infarction

Abstract

Introduction: Arterial ischemic stroke (AIS) occurs in 1-2 children/100,000/year. Most children have neurologic deficits post-AIS, but the influence of infarct volume on neurologic outcome is understudied. While acute infarct volume {best measured as a percentage of total brain volume infarcted (%aTBVi) in the growing brain} likely predicts outcome, acute scans are not always available in children. Chronic infarct volumes are underestimated by direct measurement due to contraction. A method for estimating acute infarct volumes from chronic scans is needed. We developed and compared the reliability of three methods for estimating the %aTBVi from values measured on chronic images. Methods: A retrospective IRB-approved study studying children (age 1 month-17 years) with AIS enrolled 158 patients. Those with acute (<3 days) and chronic (>90 days) MRIs were manually segmented by a pediatric neuroradiologist. Method 1 (direct method, used as control) estimated %aTBVi by measuring chronic infarct volume (cVI) and dividing by total brain volume. Method 2 (OFC method) estimated %aTBVi by subtracting the total non-infarcted brain volume from an extrapolated total brain volume based on orbitofrontal circumference (OFC). Method 3 (contraction method, Figure 1) used a regression model to apply a correction factor to the direct measurement of cVI that was then divided by total brain volume to estimate %aTBVi. Intraclass correlation compared estimated %aTBVi of the three methods to the gold standard %aTBVi calculated from manual segmentation of acute scans. Results: Inclusion criteria were met by 86 patients. The control direct method had excellent reliability (ICC 0.79), although it was exceeded by the contraction method (Figure 1, ICC=0.86), while the OFC method reliability was poor (ICC=0.42). Conclusion: %aTBVi is reliably estimated in children with AIS with only chronic imaging via the contraction method.