Virtual Monochromatic Imaging for Dual Energy Tomosynthesis: A Numerical Simulation Study
Qi Z, and Mahoney M. Virtual Monochromatic Imaging for Dual Energy Tomosynthesis: A Numerical Simulation Study. J Med Phys 2019; 46(6):e347-e348.
J Med Phys
Purpose: To develop a virtual monochromatic imaging method for Dual Energy Breast Tomosynthesis and evaluate the method via a numerical simulation study. Methods: The proposed method is composed of the following steps: first, the pair of low kVp and high kVp projection images was mapped to equivalent thicknesses of two known reference materials (LDPE and PMMA in this study); then, in the projection domain, for any arbitrary photon energy, the virtual projection data for the object is generated by summing up monochromatic projection data (at the particular energy) for each reference material; and, finally, the virtual monochromatic projection data at all views are reconstructed into tomosynthetic image slices by an analytical algorithm. The method is validated by a numerical simulation. The phantom is a water equivalent elliptical cylinder containing three spherical inserts (1 cm diameter) made of different materials, including glandular tissue, adipose and oily cyst. The simulated Tomosynthesis acquisition used two different xray beam settings: 21 kVp with 0.7 mm Al, and 39 kVp with 0.7 mm Al and 0.2 mm Cu, and, for each beam setting, collected 15 views of projection data, one view per degree. Contrast between different types of tissue on the reconstructed images are measured for different energies. Results: Virtual monochromatic images at 10 keV show markedly increased contrast for the spherical inserts compared to those at higher keVs (15 keV and 20 keV). Glandular tissue and Cyst have nearly identical signal on virtual monochromatic images at 20 keV, whereas there is 14% contrast between the two simulated tissue types at 10 keV. Conclusion: Virtual monochromatic imaging is possible with Dual Energy Tomosynthesis. It involves no use of iodinated contrast. Its benefits include increased image contrast and potential values as a quantitative tool for improved tissue differentiation.