Abstract Archives of the RSNA, 2014
SSQ02-08
Coronary Artery Calcium Quantification Based on Virtual Non-contrast Dual Energy CT Data Sets
Scientific Papers
Presented on December 4, 2014
Presented as part of SSQ02: Cardiac (Dual Energy CT)
Patricia M. Carrascosa MD, Presenter: Research Consultant, General Electric Company
Macarena De Zan, Abstract Co-Author: Nothing to Disclose
Carlos Capunay MD, Abstract Co-Author: Nothing to Disclose
Mauro Namias MSc, MEng, Abstract Co-Author: Nothing to Disclose
Alejandro Deviggiano MD, Abstract Co-Author: Nothing to Disclose
Silvina Cipriano, Abstract Co-Author: Nothing to Disclose
Gaston Rodriguez Granillo, Abstract Co-Author: Nothing to Disclose
Javier Vallejos MD, MBA, Abstract Co-Author: Nothing to Disclose
To evaluate the feasibility of using CT virtual non-contrast enhanced (VNC) series derived from dual-energy CT (DECT) imaging studies for coronary artery calcium quantification.
The study was approved by the institutional review board; all patients provided written informed consent. In 60 patients, we prospectively underwent single energy non-contrast enhanced CT calcium scoring scan at 120 kV followed by coronary CT angiography performed in dual energy mode based on rapid kilovoltage switching between 80 and 140 kV, using 0.625 mm slice thickness. Both scans were carried out on a 128-slice DECT scanner (Discovery CT750 HD; GE Medical Systems). Material basis decomposition images were created for the iodine/water pair. Polychromatic images at 120 kV were generated from the material density images by thresholding water images above 1100 mg/cm3 to suppress iodine content with dedicated software which gave virtual non-contrast images in HU as a result. Both data sets were analyzed using dedicated calcium scoring software (Smart Score 4.0; GE Medical Systems). The AJ-130, the mass and the volume scores and the percentile rank for age and gender of each patient were calculated. Correlation between quantified calcium on VNC images and true non-contrast enhanced coronary artery calcium scoring images, and the percentile rank was performed by the Spearman rank order correlation coefficient.
There was excellent correlation between AJ-130, the mass score and the volume score on VNC images and the true non-contrast enhanced series (r= 0.95 (0.93 to 0.97), 0.95 (0.93 to 0.97), and 0.96 (0.94 to 0.98), respectively; p, 0.001) globally. There was also excellent agreement between Multiethnic Study of Atheroschlerosis percentile ranks derived from predicted versus actual calcium scores and the percentile rankings, 0.88 (0.81 to 0-93).
The excellent correlation between the amounts of calcium measured on non-contrast enhanced CT calcium scoring images and VNC data derived from DECT provide an opportunity to avoid the need for a dedicated CT scan for calcium scoring.
Coronary artery calcium scoring (CCS) has been shown to increase the predictive accuracy of established multivariate risk factor models and it is an independent predictor of cardiovascular events.
Carrascosa, P,
De Zan, M,
Capunay, C,
Namias, M,
Deviggiano, A,
Cipriano, S,
Rodriguez Granillo, G,
Vallejos, J,
Coronary Artery Calcium Quantification Based on Virtual Non-contrast Dual Energy CT Data Sets. Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL.
http://archive.rsna.org/2014/14011473.html