Arterial diameters on catheter and computed tomographic angiography

Abstract

Background The diagnosis of cerebral vasospasm is hampered by lack of an accurate, noninvasive test. Computed tomographic angiography (CTA) may be useful but the correlation between arterial diameters determined from catheter digital subtraction angiography (DSA) and CTA over a range of artery sizes would need to be determined to show this. The purpose of this study was to determine the correlation between artery diameters measured on DSA and multidetector CTA. Methods Two hundred forty artery diameters were measured in DSA and CTA from 46 patients who underwent both studies within 12 hours of each other. Axial cross section, maximum intensity projection, and volume-rendered images were measured and compared by linear correlation. Two independent readers measured CTA diameters to determine interobserver variability by linear correlation. Values also were categorized and compared by χ 2 and κ statistics. Analysis was repeated with unmeasurable arteries assigned a value of 0. Results There were significant correlations between arterial diameters measured on DSA and those from CTA measured by any method ( R 2 ranging from 0.45 to 0.76, P < .0001), although there was a tendency for the slope of this relationship to be less than 1, indicating underestimation of diameter of large and overestimation of diameter of small arteries with CTA. Computed tomographic angiography diameters also correlated significantly between the 2 reviewers with higher values often when unmeasureable arteries were assigned a value of 0 ( κ = 0.23-0.55, P < .0001). Conclusion Arterial diameters measured on multidetector CTA correlate well with those determined from DSA and should permit use of CTA for quantitative study of cerebral vasospasm and other conditions requiring accurate measurement of arterial diameters. The limitation of CTA remains the inability to measure some arteries due to artifact.