Coronary calcium scoring: modelling, predicting and correcting for the effect of CT scanner spatial resolution on Agatston and volume scores.
The purpose of this study was to evaluate the impact of spatial resolution on coronary calcium scoring by x-ray CT, to assess the scoring performance of different CT scanners as they are operated in the field and to correct for the effects of CT scanner spatial resolution on coronary calcium scoring. A phantom consisting of five aluminium wires of known diameter in water was used to measure spatial resolution and to assess scoring performance. Fourteen CT scanners (three helical, two dual, two electron-beam and seven multi-detector) from four manufacturers were evaluated, some under different operating conditions. One scanner was monitored over a 3 month period and again 6 months later. Both spatial resolution and image pixel size significantly affect calcium scoring results. Spatial resolution can be measured with a precision of about 2%. Scanner spatial resolution ranged from 1 to 1.7 mm full-width-half-maximum (FWHM), and pixel size from 0.25 to 0.86 mm. Spatial resolution differences introduce systematic scoring differences that range from 38% to 1100% depending on wire size. Significant temporal variations in spatial resolution were observed in the monitored scanner. By correcting all the scanners to the same target spatial resolution, the standard deviation of individual scanners with respect to a mean value (the spread) can be reduced by 25-70% for different wires. In conclusion, scanner spatial resolution significantly affects calcium scoring and should be controlled for. Scanner performance can change over time. Under ideal conditions, CT scanners should be operated with a standard spatial resolution for calcium scoring. When this is not possible, post-processing correction is a viable alternative.