Program Information
Estimation of DgN Coefficients in Breast CT Using Heterogeneous Fibroglandular Tissue Distributions and Breast-Shaped Phantoms
A M. Hernandez1*, J M. Boone2 , (1) Department of Radiology, Biomedical Engineering Graduate Group, University of California Davis, Sacramento, CA, (2) Departments of Radiology and Biomedical Engineering, University of California Davis, Sacramento,
Presentations
TH-AB-601-4 (Thursday, August 3, 2017) 7:30 AM - 9:30 AM Room: 601
Purpose: To compare differences in normalized mean glandular dose coefficients for breast CT “DgNCT” when realistic distributions of glandular breast tissue are considered.
Methods: A large cohort of breast CT data sets were used to quantify the 3D distribution of fibroglandular tissue in the breast. Average glandular distributions were then computed and grouped in accordance with previously-reported realistic breast-shaped phantoms of various sizes (by volume). A validated MCNP6 Monte Carlo dose simulation engine was used to compute monoenergetic DgNCT(E) coefficients for photon energies ranging from 8 to 70 keV (1 keV intervals) in the breast-shaped phantoms composed of either the realistic heterogeneous distribution of glandular tissue “DgNCTʰᵉᵗᵉʳᵒ(E)” or the historical assumption of a homogeneous mixture. Polyenergetic dose coefficients (pDgNCT) were also determined by spectrally weighting the DgNCT(E) coefficients using the TASMICS-BCT spectral model. Glandular dose comparisons were made for the two different breast compositions in the median-sized phantom.
Results: The monoenergetic DgNCTʰᵉᵗᵉʳᵒ(E) coefficients were consistently lower than those estimated using the homogeneous approximation and these differences decreased with increasing photon energy. The pDgNCTʰᵉᵗᵉʳᵒ coefficients were on average 25.2% and 12.1% lower than the pDgNCTʰᵒᵐᵒ coefficients for 1.5 mm of Al and 0.2 mm of Cu filtration, respectively, averaged across tube potentials from 35 to 70kV.
Conclusion: This preliminary work demonstrates that glandular dose in breast CT is lower than previously assumed using a simple homogeneous breast tissue approximation. This is due to an overestimation of the glandular dose at the surface of the breast, where the dose deposition is exponentially higher. Future work will focus on investigating the dependence of pDgNCT on phantom size, volumetric glandular fraction, location of the glandular distribution within the breast, and x-ray technique factors.
Funding Support, Disclosures, and Conflict of Interest: This communication was funded in part by NIH grants P30 CA093373 and R01 CA181081. Comments made are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Contact Email: