Adaptive statistical iterative reconstruction algorithm for measurement of vascular diameter on computed tomographic angiography in vitro.

OBJECTIVES: To evaluate the accuracy of vascular diameter measurement on computed tomographic (CT) angiography using adaptive statistical iterative reconstruction (ASIR). METHODS: We scanned 9 vascular models with 3 wall thicknesses and filled with 3 densities of contrast material using a 64-detector CT unit, reconstructed images using ASIR (0%, 20%, 40%, 60%, 80%, and 100%), and repeated 20 separate diameter measurements for each model. We evaluated the distribution of image noise for the 0% and 100% ASIR. RESULTS: For all vascular models, measurement errors differed significantly (P < 0.0001) among the percentages of ASIR, tending to increase as the percentage of ASIR increased for models filled with 246 and 354 Hounsfield units of contrast medium. The degree of image noise depended on the substance within the model with 100% ASIR. CONCLUSIONS: Adaptive statistical iterative reconstruction can enhance errors in diameter measurement on CT angiography and should be applied carefully to evaluate small vessels.