Bowtie filters for dedicated breast CT: theory and computational implementation.

Area of Science:

• Medical Physics
• Radiological Imaging
• Biomedical Engineering

Background:

• Dedicated breast computed tomography (CT) requires specialized imaging components to optimize image quality and minimize patient radiation exposure.
• Bowtie filters are crucial in CT systems to shape the X-ray spectrum, reducing beam hardening artifacts and improving dose efficiency.

Purpose of the Study:

• To design and evaluate novel bowtie filters for dedicated breast CT with the primary goals of improving image quality and reducing patient dose.
• To explore different design strategies for bowtie filters, including spectral matching and basis material decomposition, to achieve superior performance.

Main Methods:

• Three distinct bowtie filter designs were developed using analytical computational methods and linear attenuation coefficients for a 14-cm diameter breast phantom.
• Designs focused on spectral matching, basis material decomposition, and elimination of beam hardening effects.
• Monte Carlo simulations and analysis of reconstructed images were employed to evaluate dose distribution, noise uniformity, and contrast-to-noise ratio (CNR) homogeneity.

Main Results:

• All three bowtie filter designs reduced the dynamic range requirements on the detector compared to systems without filters.
• Simulations demonstrated a significant reduction in peripheral breast dose (up to 61%) and improved uniformity in noise and CNR distributions.
• The validated design concepts were extended to a 3D anthropomorphic bowtie filter for comprehensive breast imaging.

Conclusions:

• The developed bowtie filters are robust against variations in breast size, composition, and X-ray tube voltage.
• These filters offer a substantial improvement in patient dose reduction and image quality for dedicated breast CT applications.
• The study successfully demonstrated the efficacy of computational and simulation techniques in designing advanced bowtie filters for medical imaging.