RESUMO
RATIONALE AND OBJECTIVES: Quantitatively analyze the computed tomography (CT) attenuation effects caused by bismuth shields, which are used to reduce superficial organ dose. MATERIALS AND METHODS: The solid water uniformity section of the American College of Radiology CT phantom was scanned with a modified chest CT protocol. Scans were performed with a bismuth breast shield in multiple configurations, emphasizing three clinically relevant orientations. Attenuation effects were measured as changes in mean Hounsfield unit (HU) values of equal midsagittal regions of interest (ROI). Multiple statistical techniques were used in regression analysis. RESULTS: Bismuth shielding resulted in significant positive shifts of the expected Hounsfield unit values. The mean nonshielded CT attenuation was -0.16 ± 0.75 HU. Based on the clinically relevant ROI distance from the shield (~3-16 cm), the shielded values ranged from 43.8-4 HU, 45.8-10.1 HU, and 50.6-4.5 HU for shields 1, 2, and 3, respectively. All shield configurations displayed a statistically significant shift (P < .0001) at all distance ranges. The best fitting regression model was a quadratic function of distance versus logarithmic function of HU. A prediction table of the approximate shift in water HU values as a function of ROI distance from the shield was generated per shield type from their respective close-fitting regressions. CONCLUSIONS: The data support the claim that bismuth shields increase the attenuation of water, which can cause inaccurate characterization of simple fluid, giving the appearance of complex fluid or even solid density. However, there is potential for anticipation of the attenuation effects to validate continued use of these shields for dose reduction.