Program Information
Dual-Energy Calibration for Breast Density Measurement Using Spectral Mammography
H Ding*, H Cho , N Kumar , D Sennung , S Molloi , Department of Radiological Sciences, University of California, Irvine, CA
Presentations
SU-D-204-1 (Sunday, July 12, 2015) 2:05 PM - 3:00 PM Room: 204
Purpose: To investigate the feasibility of minimizing the systematic errors in dual-energy breast density quantification induced by the use of tissue-equivalent plastic phantoms as the calibration basis materials.
Methods: Dual-energy calibration using tissue-equivalent plastic phantoms was performed on a spectral mammography system based on scanning multi-slit Si strip photon-counting detectors. The plastic phantom calibration used plastic water and adipose-equivalent phantoms as the basis materials, which have different x-ray attenuation properties compared to water and lipid in actual breast tissue. Two methods were used to convert the dual-energy decomposition measurements in plastic phantom thicknesses into true water and lipid basis. The first method was based entirely on the theoretical x-ray attenuation coefficients of the investigated materials in the mammographic energy range. The conversion matrix was determined from least-squares fitting of the target material using the reported attenuation coefficients of water and lipid. The second method was developed based on experimental calibrations, which measured the low-and high-energy signals of pure water and lipid of known thicknesses. A non-linear rational function was used to correlate the decomposed thicknesses to the known values, so that the conversion coefficients can be determined. Both methods were validated using independent measurements of water and lipid mixture phantoms. The correlation of the dual-energy decomposition measurements and the known values were studied with linear regression analysis.
Results: There was an excellent linear correlation between the converted water thicknesses and the known values. The slopes of the linear fits were determined to be 0.63 and 1.03 for the simulation and experimental results, respectively. The non-linear fitting in the experimental approach reduced the root-mean-square (RMS) errors from approximately 3.4 mm to 1.5 mm.
Conclusion: The results suggested that conversion of the dual-energy measurements into water and lipid thicknesses minimized the systematic errors in tissue decomposition studies.
Contact Email: