Evaluation of the BreastSimulator software platform for breast tomography.

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Computational Modeling

Background:

• Developing advanced breast imaging techniques is crucial for early cancer detection.
• Computed tomography (CT) scanners dedicated to breast imaging offer potential for improved diagnostic accuracy.
• Accurate digital breast models are needed for simulating and optimizing these specialized CT systems.

Purpose of the Study:

• To evaluate BreastSimulator software for creating 3D digital breast models.
• To assess its utility in simulating and optimizing dedicated breast CT scanners.
• To validate the generated breast phantoms against clinical data.

Main Methods:

• Eight 3D digital breast phantoms with varying glandular fractions (10%-35%) were created using BreastSimulator.
• X-ray CT projections were simulated using cone-beam CT and reconstructed with the FDK algorithm.
• Simulations utilized various mono-energetic and poly-energetic X-ray spectra, with validation against clinical breast CT images.

Main Results:

• The software-generated breast phantoms accurately reproduced realistic anatomical features.
• Quantitative analysis using the power-law exponent (β) showed close agreement between simulated and clinical breast CT data.
• The β coefficients derived from simulated images validated the anatomical accuracy of the digital breast models.

Conclusions:

• BreastSimulator is a valuable tool for generating realistic 3D digital breast models for breast CT imaging research.
• The software facilitates the simulation and optimization of dedicated breast CT scanners.
• This research tool can significantly contribute to the advancement and refinement of breast CT imaging techniques.