A method for coronary artery calcium scoring using contrast-enhanced computed tomography

BACKGROUND: Limitations to the coronary calcium score include its requirement for noncontrast imaging and radiation exposure that approaches current methods for contrast-enhanced CT angiography. OBJECTIVES: We sought to derive and validate a method of measuring the coronary artery calcium score (CACS) from standard contrast-enhanced CT, obviating the need for a second non-contrast calcium scan. METHODS: The volume of intramural calcium of >320 HU in major coronary vessels was measured in 90 contrast-enhanced and traditional non-contrast calcium scan pairs. An empiric conversion factor was derived to convert the small voxel contrast-enhanced calcium volume to an Agatston calcium score. The accuracy of this technique was then prospectively validated in 120 consecutive patients undergoing clinical calcium scans and contrasted-enhanced coronary CT. Eleven patients were excluded from analysis because of the prespecified criteria of excessive noise in the contrast-enhanced CT or total coronary artery occlusion. RESULTS: The Pearson correlation of the contrast scan-derived calcium score with the measured CACS was r(2) = 0.99. With standard CACS risk bands, agreement of the contrast-enhanced calcium score estimate with the measured CAC by quadratic weighted kappa was 0.96. The 95% limits of agreement (Agatston units) were given by +/-(3.2+0.14 x CACS +4.44 root CACS). Inter-observer and intra-observer reliability with the intraclass correlation was 0.99. CONCLUSION: The calcium score can be accurately measured from contrast-enhanced cardiac CT scans with the use of a Hounsfield unit threshold of 320. (C) 2012 Published by Elsevier Inc. on behalf of Society of Cardiovascular Computed Tomography.