부산대학교 도서관

Introduction:Proficiency required to execute CT perfusion (CTP) protocols is a limiting factor in its use in acute stroke. We propose to calculate perfusion parametric maps and measure ischemic volumes using readily available non-contrast CT (NCCT) and multiphase CT angiography (mCTA) images.Materials and Methods:Twenty-five patients presenting with acute ischemic stroke were included in this study. Our proposed dynamic sequence (multiphase CT angiography-perfusion, mCTA-P) consisted of the NCCT as the pre-contrast baseline and three phases of mCTA, which corresponded to the peak arterial, peak venous, and late venous phases at 8 s intervals. CTP was acquired after mCTA and consisted of 22 dynamic images acquired over 60 s at 2.8 s intervals. A prototype model-based deconvolution algorithm (CT Perfusion 4D, GE Healthcare) was used to calculate cerebral blood flow (CBF) and Tmax maps for each series. Infarct was classified as voxels that satisfied both a time-dependent relative CBF threshold and Tmax > 8 s while penumbral voxels satisfied either threshold but not both.Results:Median (interquartile range) 24-hour follow-up infarct volume was 18.6 (4.7 to 34.3) ml and median stroke onset-to-CTP time was 124.0 (70.5 to 201.5) min. Bland-Altman analysis revealed good agreement between CTP and mCTA-P volume measurements as mean differences (limits of agreement) were -1.0 (-14.9 to 12.9) ml for infarct and 8.4 (-42.4 to 59.1) ml for penumbra. Intraclass correlation (95% confidence interval, p < 0.05) between CTP and mCTA-P volumes were 0.72 (0.46 to 0.87) for infarct and 0.68 (0.41 to 0.85) for penumbra, indicating good to moderate reliability.Conclusion:Quantitative perfusion can be estimated from NCCT and mCTA without introducing additional scan time, radiation dose, and contrast injections associated with CTP. Our technique allows assessments of early ischemic changes and collaterals to be augmented with quantitative perfusion measurements of ischemic volumes.