A method for benchmarking CT scanners.

ABSTRACT

This study involved the development of an objective method to compare the performance of five CT scanners for the purpose of benchmarking. The method used to assess the scanners was to determine the dose-normalised noise at a spatial resolution of 5.5 cm(-1). This gave a dose-normalised percent noise between 0.37% and 0.76%. The scanners were also assessed for radiation dose to patients undergoing abdomen and head CT examinations. Patients' dose-length product (DLP) for the abdomen clinical examinations varied from 305 to 685 mGy-cm, and for the head clinical examinations from 333 to 900 mGy-cm. The study results demonstrated that the comparison of dose and spatial resolution normalised percent noise levels is a useful method of comparing CT scanner performance.