Breast Shielding Combined With an Optimized Computed Tomography Pulmonary Angiography Pregnancy Protocol: A Special Use-Case for Shielding?

ABSTRACT

OBJECTIVES: To determine the impact of breast shields on breast dose and image quality when combined with a low-dose computed tomography pulmonary angiography (CTPA) protocol for pregnancy. METHODS: A low-dose CTPA protocol, with and without breast shields, was evaluated by anthropomorphic phantom and 20 prospectively recruited pregnant participants from January to October 2019. Thermoluminescent dosimeters measured surface and absorbed breast dose in the phantom and surface breast dose in participants. The Monte-Carlo method estimated the absorbed breast dose in participants. Image quality was assessed quantitatively by regions of interest analysis and subjectively by the Likert scale. Doses and image quality for CTPA alone were compared with CTPA with breast shields. RESULTS: Mean surface and absorbed breast dose for CTPA alone were 1.3±0.4 and 2.8±1.5 mGy in participants, and 1.5±0.7 and 1.6±0.6 mGy in the phantom. Shielding reduced surface breast dose to 0.5±0.3 and 0.7±0.2 mGy in the phantom (66%) and study participants (48%), respectively. Absorbed breast dose reduced to 0.9±0.5 mGy (46%) in the phantom.Noise increased with breast shields at lower kV settings (80 to 100 kV) in the phantom; however, in study participants there was no significant difference between shield and no-shield groups for main pulmonary artery noise (no-shield 34±9.8, shield 36.3±7.2, P =0.56), SNR (no-shield 11.2±3.7, shield 10.8±2.6, P =0.74) or contrast-to-noise ratio (no-shield 10.0±3.3, shield 9.3±2.4, P =0.6). Median subjective image quality scores were comparable (no-shield 4.0, interquartile range 3.5 to 4.4, shield 4.3, interquartile range 4.0 to 4.5). CONCLUSION: Combining low-dose CTPA with breast shields confers additional breast-dose savings without impacting image quality and yields breast doses approaching those of low-dose scintigraphy, suggesting breast shields play a role in protocol optimization for select groups.