Electron beam CT versus 16-slice spiral CT: how accurately can we measure coronary artery calcium volume?

The purpose of this study was to investigate how accurately CAC volume we can be measured using electron beam computed tomography (CT) and 16-slice spiral CT. CAC models with known volume attached to a cardiac phantom were scanned. The error of measurement, variability between measured and real volumes, and inter-scan measurement variability were obtained. For spiral CT, seven different parameters were included: (1) slice thickness (0.625 mm, 1.25 mm and 2.5 mm), (2) retrospective spiral electrocardiograph (ECG)-gated or prospective axial ECG-triggering, (3) overlapping or non-overlapping. The error of measurement was 15% on electron beam CT and 8-20% on spiral CT. CAC volumes were underestimated in 92% and overestimated in 8% of the electron beam CT scans. Volumes were underestimated in 79%, correct in 5% and overestimated in 16% of the spiral CT scans. The best measurement and the least variability was observed on 0.625-mm retrospective spiral ECG-gated CT (error of 8%), a significant result (t-test: P<0.01) when compared with electron beam CT. CAC volume measurement on CT scanners may be significantly different and often underestimates the real volume of CAC. For precise evaluation of CAC volume, thin-slice retrospective spiral ECG-gated scan using a spiral CT scanner is desirable.