skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Characterization of scatter in digital mammography from use of Monte Carlo simulations and comparison to physical measurements

Abstract

Purpose: Monte Carlo simulations were performed with the goal of verifying previously published physical measurements characterizing scatter as a function of apparent thickness. A secondary goal was to provide a way of determining what effect tissue glandularity might have on the scatter characteristics of breast tissue. The overall reason for characterizing mammography scatter in this research is the application of these data to an image processing-based scatter-correction program. Methods: MCNPX was used to simulate scatter from an infinitesimal pencil beam using typical mammography geometries and techniques. The spreading of the pencil beam was characterized by two parameters: mean radial extent (MRE) and scatter fraction (SF). The SF and MRE were found as functions of target, filter, tube potential, phantom thickness, and the presence or absence of a grid. The SF was determined by separating scatter and primary by the angle of incidence on the detector, then finding the ratio of the measured scatter to the total number of detected events. The accuracy of the MRE was determined by placing ring-shaped tallies around the impulse and fitting those data to the point-spread function (PSF) equation using the value for MRE derived from the physical measurements. The goodness-of-fit was determined for eachmore » data set as a means of assessing the accuracy of the physical MRE data. The effect of breast glandularity on the SF, MRE, and apparent tissue thickness was also considered for a limited number of techniques. Results: The agreement between the physical measurements and the results of the Monte Carlo simulations was assessed. With a grid, the SFs ranged from 0.065 to 0.089, with absolute differences between the measured and simulated SFs averaging 0.02. Without a grid, the range was 0.28–0.51, with absolute differences averaging −0.01. The goodness-of-fit values comparing the Monte Carlo data to the PSF from the physical measurements ranged from 0.96 to 1.00 with a grid and 0.65 to 0.86 without a grid. Analysis of the data suggested that the nongrid data could be better described by a biexponential function than the single exponential used here. The simulations assessing the effect of breast composition on SF and MRE showed only a slight impact on these quantities. When compared to a mix of 50% glandular/50% adipose tissue, the impact of substituting adipose or glandular breast compositions on the apparent thickness of the tissue was about 5%. Conclusions: The findings show agreement between the physical measurements published previously and the Monte Carlo simulations presented here; the resulting data can therefore be used more confidently for an application such as image processing-based scatter correction. The findings also suggest that breast composition does not have a major impact on the scatter characteristics of breast tissue. Application of the scatter data to the development of a scatter-correction software program can be simplified by ignoring the variations in density among breast tissues.« less

Authors:
 [1];  [2]
  1. Department of Diagnostic and Interventional Imaging, University of Texas Medical School at Houston, Houston, Texas 77030 (United States)
  2. Department of Radiology, University of Florida, Gainesville, Florida 32610 (United States)
Publication Date:
OSTI Identifier:
22320352
Resource Type:
Journal Article
Journal Name:
Medical Physics
Additional Journal Information:
Journal Volume: 41; Journal Issue: 11; Other Information: (c) 2014 American Association of Physicists in Medicine; Country of input: International Atomic Energy Agency (IAEA); Journal ID: ISSN 0094-2405
Country of Publication:
United States
Language:
English
Subject:
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND TECHNOLOGY; 60 APPLIED LIFE SCIENCES; ACCURACY; ADIPOSE TISSUE; BEAMS; BIOMEDICAL RADIOGRAPHY; COMPARATIVE EVALUATIONS; COMPUTERIZED SIMULATION; DENSITY; ECONOMICS; EQUATIONS; FILTERS; IMAGE PROCESSING; MAMMARY GLANDS; MONTE CARLO METHOD; PHANTOMS; PLANT TISSUES; PULSES; THICKNESS; VARIATIONS

Citation Formats

Leon, Stephanie M., E-mail: Stephanie.Leon@uth.tmc.edu, Wagner, Louis K., and Brateman, Libby F. Characterization of scatter in digital mammography from use of Monte Carlo simulations and comparison to physical measurements. United States: N. p., 2014. Web. doi:10.1118/1.4894808.
Leon, Stephanie M., E-mail: Stephanie.Leon@uth.tmc.edu, Wagner, Louis K., & Brateman, Libby F. Characterization of scatter in digital mammography from use of Monte Carlo simulations and comparison to physical measurements. United States. https://doi.org/10.1118/1.4894808
Leon, Stephanie M., E-mail: Stephanie.Leon@uth.tmc.edu, Wagner, Louis K., and Brateman, Libby F. 2014. "Characterization of scatter in digital mammography from use of Monte Carlo simulations and comparison to physical measurements". United States. https://doi.org/10.1118/1.4894808.
@article{osti_22320352,
title = {Characterization of scatter in digital mammography from use of Monte Carlo simulations and comparison to physical measurements},
author = {Leon, Stephanie M., E-mail: Stephanie.Leon@uth.tmc.edu and Wagner, Louis K. and Brateman, Libby F.},
abstractNote = {Purpose: Monte Carlo simulations were performed with the goal of verifying previously published physical measurements characterizing scatter as a function of apparent thickness. A secondary goal was to provide a way of determining what effect tissue glandularity might have on the scatter characteristics of breast tissue. The overall reason for characterizing mammography scatter in this research is the application of these data to an image processing-based scatter-correction program. Methods: MCNPX was used to simulate scatter from an infinitesimal pencil beam using typical mammography geometries and techniques. The spreading of the pencil beam was characterized by two parameters: mean radial extent (MRE) and scatter fraction (SF). The SF and MRE were found as functions of target, filter, tube potential, phantom thickness, and the presence or absence of a grid. The SF was determined by separating scatter and primary by the angle of incidence on the detector, then finding the ratio of the measured scatter to the total number of detected events. The accuracy of the MRE was determined by placing ring-shaped tallies around the impulse and fitting those data to the point-spread function (PSF) equation using the value for MRE derived from the physical measurements. The goodness-of-fit was determined for each data set as a means of assessing the accuracy of the physical MRE data. The effect of breast glandularity on the SF, MRE, and apparent tissue thickness was also considered for a limited number of techniques. Results: The agreement between the physical measurements and the results of the Monte Carlo simulations was assessed. With a grid, the SFs ranged from 0.065 to 0.089, with absolute differences between the measured and simulated SFs averaging 0.02. Without a grid, the range was 0.28–0.51, with absolute differences averaging −0.01. The goodness-of-fit values comparing the Monte Carlo data to the PSF from the physical measurements ranged from 0.96 to 1.00 with a grid and 0.65 to 0.86 without a grid. Analysis of the data suggested that the nongrid data could be better described by a biexponential function than the single exponential used here. The simulations assessing the effect of breast composition on SF and MRE showed only a slight impact on these quantities. When compared to a mix of 50% glandular/50% adipose tissue, the impact of substituting adipose or glandular breast compositions on the apparent thickness of the tissue was about 5%. Conclusions: The findings show agreement between the physical measurements published previously and the Monte Carlo simulations presented here; the resulting data can therefore be used more confidently for an application such as image processing-based scatter correction. The findings also suggest that breast composition does not have a major impact on the scatter characteristics of breast tissue. Application of the scatter data to the development of a scatter-correction software program can be simplified by ignoring the variations in density among breast tissues.},
doi = {10.1118/1.4894808},
url = {https://www.osti.gov/biblio/22320352}, journal = {Medical Physics},
issn = {0094-2405},
number = 11,
volume = 41,
place = {United States},
year = {Sat Nov 01 00:00:00 EDT 2014},
month = {Sat Nov 01 00:00:00 EDT 2014}
}