Comparison of three generations of electron beam tomography on image noise and reproducibility, a phantom study

Abstract

The detection of coronary artery calcification (CAC) using the electron beam tomography (EBT) scanner provides a noninvasive indicator for coronary artery disease (CAD). Physicians interested in preventative medicine also are using this modality to track atherosclerosis over time. Two new iterations of the EBT scanner have been introduced. We sought to evaluate the image quality of each machine to examine whether patients scanned on a previous model would have similar image quality and results to those scanned on the newest scanner.

Methods: This study used the C-150 XP, C300, and the e-Speed EBT scanners in high-resolution volume mode. A cork chest phantom was constructed for use as a human chest. A mixture of calcium phosphate, cornstarch and glue was placed inside the wells to simulate coronary calcium. The foci masses were 3, 5, 7, 10, 15, 20, 40, 60, 80, 100, and 200 mg (calcium mass), which provided the 55 foci of different masses and densities to simulate coronary calcium in the chest phantom. Each phantom was scanned multiple times, using both 1.5- and 3-mm slice thickness and table collimation settings with each scanner.

Results: There were no statistical differences found between the 1.5-mm and 3.0-mm slice thickness calcium foci scores (Agatston & volumetric) for all 3 EBT scanners. The C-150 XP scanner had a variability of 6.01% between 1.5-mm and 3.0-mm slice thickness. Analysis by t test revealed that the mean noise value of C-150 XP was significantly higher than the C300, e-Speed (50 milliseconds), and e-Speed (100 milliseconds) with P values of 0.001, 0.025, and 0.001, respectively. Comparison of 1.5-mm versus 3.0-mm slice thickness noise value showed a significant difference only for the C-150 XP scanner (P < 0.05).

Conclusions: The use of the 3 EBT scanners in longitudinal studies of patients coronary calcium score is feasible to obtain similar calcium score values. The C-150 XP has the greatest noise effect in comparison to the C300 and e-Speed scanners. Improved image noise should improve reproducibility of the calcium measurement with these newer devices.