Estimation of Breast Density from DICOM Header Data using a Geant4 Monte-Carlo Simulation of the Mammography Imaging Process

Type of content
Theses / Dissertations
Publisher's DOI/URI
Thesis discipline
Medical Physics
Degree name
Master of Science
Publisher
University of Canterbury
Journal Title
Journal ISSN
Volume Title
Language
English
Date
2017
Authors
Cunningham, Brett
Abstract

Breast density is the fractional measurement of glandular tissue versus that of adipose and connective tissue in the breast. This value is often determined from the subjective interpretation of a mammography image by a radiologist. Because this may vary significantly between interpretations, a more formal measure is required to describe a breast density in terms of it's radiographic characteristics, and can be clearly understood and communicated. Geant4 Monte-Carlo package was used to develop a simulation

Geant4 Monte-Carlo package was used to develop a simulation of the mammography imaging process that included relevant geometries specific to a GE Senograph Essential mammography unit. Filter transmission data was collected for filter materials molybdenum, rhodium and aluminium with anode materials molybdenum, rhodium and tungsten over energy ranges 24- 36kV with 2kV increments. Subsequent breast transmission data was col- lected for selected beam parameters using polymethyl methacrylate(PMMA) and water as breast materials.

From this transmission data a water equivalent thickness was able to be calculated with the aid of Dance, Skinner, Young, Beckett, and Kotre (2000) equivalent breast thickness calculations, as well as an associated breast den- sity, or glandularity. It was found that the mean simulated breast trans- mission uncertainty was ±4.8%, and mean GE Senograph Essential breast transmission uncertainty was ±5.2%. This meant the majority of trans- mission curves fell within each others error bars as the mean transmission difference was 8.6%. Ultimately a water equivalent thickness was calculated to within a mean accuracy of ±7.0%.

Description
Citation
Keywords
Ngā upoko tukutuku/Māori subject headings
ANZSRC fields of research
Rights
All Rights Reserved