Abstract 1601: Acute Pulmonary Embolism: The Optimal Scan Delay for Contrast Medium Injection in Spiral CT Angiography
AUTHORS:
I.J. Hartmann, MD, R.T. Lo, MD, W. De Monye, MD, P.M. Pattynama, MD, PhD, P.F. Van Waes, MD, PhD.
ABSTRACT:
PURPOSE: For optimal image quality of spiral CT of pulmonary arteries maximal vascular enhancement is required. The aim of this study was to determine the optimal scan delay in spiral CT angiography (SCTA) and to assess whether easily obtainable clinical parameters can be used to predict optimal scan delay.
METHOD AND MATERIALS: In 135 patients SCTA was performed with 5 mm collimation, 1.0 pitch, reconstruction index of 2 mm, and 0.9 mg iodium/s i.v. injection. In 83 patients we measured the transit time of a test bolus with a ROI set at the left ventricle (LV) and descending aorta. The scan delay was individualized according to the initial rise of the time-density curve of the LV or, when this was not available, the aorta. In the 52 remaining other patients we used a standard, fixed delay of 20 s. We assessed the vessel enhancement up to the level of segmental branching.
RESULTS: Test bolus injection yielded 70/83 interpretable time-density curves. The quality of vessel enhancement was good in 87%, moderate in 3% and poor in 10%. The results with the 20 s fixed scan delay were not significantly different: good quality in 79%, moderate in 8% and poor in 13%. The mean contrast arrival time was 13.0 s (range 5-33), longer in men than in women (15.0 s vs 12.0 s, p=0.02). The arrival time increased with age (p=0.03), but showed no relation with heart rate, blood pressure, body length, body weight or body surface. In 4 patients the contrast arrival time exceeded the upper limit of the 95%-confidence interval of 22.6 s, but in only 2 of these patients this resulted in poor quality vessel enhancement.
CONCLUSIONS: Although individual contrast delay times vary from 5 to 33 s, the use of a fixed scan delay of 20 s is suitable. Individual contrast timing does not result in better image quality. The scan delay cannot be realibly predicted from easily obtainable clinical parameters.
