Comparative analysis of patient scatter to phantom scatter for computed tomography systems

Current computed tomography (CT) shielding practices are largely based on calculations of scattered radiation emitted from an acrylic head or body phantom, such as the Computed Tomography Dose Index (CTDI) phantom, or anthropomorphic phantoms of these two anatomical categories. This report considers the difference in scattered air kerma or dose from phantom models, to actual patient scatter under a variety of clinical scan conditions. Empirical patient scatter measurements recorded at different positions around the gantry, for 3 different CT scanners, resulted in average patient scatter fractions per unit Dose Length Product (DLP: mGy cm) at 1 m from the isocentre of 0.09 ± 0.03 and 0.17 ± 0.04 µGy (mGy cm)^-1 for head and body scans respectively. For the purposes of shielding design and scattered dose estimates to staff it is recommended that a single standard deviation be applied to these averages in the continued interest of conservatism. These values are reasonably comparative to the widely published scatter fractions by the National Council of Radiological Protection using the CTDI phantom, and the British Institute of Radiology using head and body anthropomorphic phantoms.