Evaluation of attenuation-based tube current control in coronary artery calcium scoring on prospective ECG-triggered 64-detector CT.

RATIONALE AND
OBJECTIVES: The aims of this study were to investigate image noise (standard deviation of computed tomographic value) and to assess variability in repeated coronary artery calcium (CAC) scoring on prospective electrocardiographically triggered 64-detctor computed tomography.
MATERIALS AND METHODS: Patients (n = 428) suspected of having coronary artery disease were scanned twice using three protocols: with tube current modified by body mass index (BMI; group A), by BMI and body height (group B), and by attenuation at the maximal heart diameter (group C). Image noise was plotted against BMI. Interscan variability of CAC scores was determined. The effective dose was estimated by computed tomographic dose index.
RESULTS: The mean effective dose and image noise, respectively, were 0.9 +/- 0.2 mSv (range, 0.6-1.5 mSv) and 19 +/- 4 Hounsfield units (HU) (range, 10-32 HU) for group A; 0.8 +/- 0.2 mSv (range, 0.5-1.4 mSv) and 18 +/- 4 HU (range, 10-31 HU) for group B; and 0.8 +/- 0.4 mSv (range, 0.3-2.2 mSv) and 20 +/- 2 HU (range, 16-26 HU) for group C. Group C used a wide dose range and controlled noise within a small range. The positive slopes of image noise versus BMI, 0.81 HU/(kg/m(2)) in group A and 0.62 HU/(kg/m(2)) in group B, suggested insufficient control of the tube current. In contrast, the nearly flat slope in group C, 0.091 HU/(kg/m(2)), indicated optimal control. The interscan variability for Agatston score, volume, and mass in patients with CAC (n = 300) was 13% (median, 8%), 12% (median, 7%), and 11% (median, 6%), respectively.
CONCLUSIONS: CAC scoring on prospective electrocardiographically triggered 64-detector computed tomography using attenuation-based tube current control has the potential to favorably control image noise with low dose and low interscan variability.