Contrast agent comparison for three-dimensional micro-CT angiography: A cadaveric study

Abstract

Barium sulfate and lead oxide contrast media are frequently used for cadaver-based angiography studies. These contrast media have not previously been compared to determine which is optimal for the visualisation and measurement of blood vessels. In this study, the lower limb vessels of 16 embalmed Wistar rats, and four sets of cannulae of known diameter, were injected with one of three different contrast agents (barium sulfate and resin, barium sulfate and gelatin, and lead oxide combined with milk powder). All were then scanned using micro-computed tomography (CT) angiography and 3-D reconstructions generated. The number of branching generations of the rat lower limb vessels were counted and compared between the contrast agents using ANOVA. The diameter of the contrast-filled cannulae, were measured and used to calculate the accuracy of the measurements by comparing the bias and variance of the estimates. Intra- and inter-observer reliability were calculated using intra-class correlation coefficients. There was no significant difference (mean difference [MD] 0.05; MD 95% confidence interval [CI] -0.83 to 0.93) between the number of branching generations for barium sulfate-resin and lead oxide-milk powder. Barium sulfate-resin demonstrated less bias and less variance of the estimates (MD 0.03; standard deviation [SD] 1.96 mm) compared to lead oxide-milk powder (MD 0.11; SD 1.96 mm) for measurements of contrast-filled cannulae scanned at high resolution. Barium sulfate-resin proved to be more accurate than lead oxide-milk powder for high resolution micro-CT scans and is preferred due to its non-toxicity. This technique could be applied to any embalmed specimen model.