Songling Liu MD, Presenter: Nothing to Disclose

Max Wintermark MD, Abstract Co-Author: Nothing to Disclose

William P. Dillon MD, Abstract Co-Author: Nothing to Disclose

To determine the reproducibility of perfusion-CT (PCT) derived measurements of cerebral blood flow (CBF), volume (CBV) and mean transit time (MTT) values in patients with suspected acute stroke.

Ten patients with suspected acute stroke underwent a CT survey including two separate PCT series(20mm thick; 2X10mm). The second 10mm slice of the first PCT series and the first 10mm slice of the second PCT series were overlapped, allowing comparison of perfusion parameters in the same patient/slice location. The acquisition parameters (gantry rotation 1 sec, 80 kVp, 100 mAs) were the same for the two PCT series, as were the amounts of iodinated contrast (4cc/sec for a total of 40 cc) and the post-processing algorithm (same location of ROI for arterial input and venous output functions, selected in the overlapping slices). The absolute values of CBF, CBV, and MTT for the whole brain, the basal ganglia, the middle cerebral artery territory and the white matter were measured on the overlapping slices using the same template of regions of interests (ROIs). CBF, CBV and MTT values were compared using paired t-tests; the coefficients of variation of CBF were calculated in order to assess reproducibility. Measurements were made including and excluding large vessels by applying a maximal enhancement threshold.

Whole brain CBF values were not significantly different on the overlapping slices (p=0.539) when pixels containing large vessels were excluded. When the pixels containing large vessels were included, the CBF values derived from the second PCT series were less than those derived from the first PCT series (p=0.003). The coefficient of variation for CBF values was 5.80% when large vessels were excluded, and 9.35% when large vessels were included in the measurement ROIs.

When large vessels are excluded from analysis, PCT parameters (CBF) are reproducible with a coefficients of variation of 5.80%. Thus, perfusion-CT can be considered as a suitable monitoring tool in longitudinal treatment trials.

Liu, S, Wintermark, M, Dillon, W, Reproducibility of Cerebral Blood Flow Measurements Using Dynamic CT Perfusion with Deconvolution.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4409268.html