Attenuation-based dynamic CT beam-shaping filtration in dependence of fan and projection angle: Evaluation of a new method for radiation exposure reduction by Monte-Carlo simulation of spatial dose distribution

Purpose: Evaluation of the potential for radiation exposure reduction of replacing static computed tomography (CT) beam-shaping filter geometry by a filter geometry dynamically adapted to the attenuation of objects of elliptic cross-section in dependence of fan and projection angle. Materials and Methods: The static geometrical shape of the bowtie filter of a CT system has been determined from primary CT X-ray spectra (120 kV_p) measured at different fan angles (0° to 21°) using a Compton spectrometer. In addition, a theoretical model of a beam-shaping filter that can be dynamically adapted to the attenuation of subjects of elliptic cross-section in function of fan and projection angle has been developed. Both bowtie models were implemented in a Geant4 Monte-Carlo simulation of CT, and spatial distributions of dose deposited in a water phantom of elliptic cross-section (16.0 cm semimajor, 12.5 cm semiminor axes) were simulated a.p. to lateral every 15° (10⁷ photons/projection). Exploiting phantom symmetry, comprehensive spatial dose distributions within the phantom were reconstructed and used to evaluate relative radiation exposure reduction achievable by replacing static with dynamic beam-shaping in CT. Results: Dynamic filtration concentrates projection-angle dependent dose deposition around the center of the fan thus minimizing contributions to more peripheral voxels, leading to relative patient skin dose reductions of ~25% a.p. or ~40% lateral. Compared to a CT acquisition without any beam-shaping filter, measured static and proposed dynamic bowtie filters allow reducing total dose deposition by 33% and 48%, respectively. Thus, an additional radiation exposure reduction by about 15% can be achieved with a filter dynamically adapted to subject shape. Conclusion: In contrast to static beam-shaping filtration used in multidetector-row computed tomography (MDCT) to date, attenuation-based dynamic filter concepts could al- ow for a significant additional reduction of radiation exposure in MDCT.