Type:
Educational Exhibit
Keywords:
Breast, Mammography, Computer Applications-General, Cancer
Authors:
G. Gennaro1, R. Highnam2; 1Padova/IT, 2Wellington/NZ
DOI:
10.1594/ecr2013/C-1033
Background
Mammography (MG) is used for the screening and diagnosis of breast cancer.
However,
breasts can have a wide range of appearance on mammography,
associated with differences in tissue composition.
In terms of X-ray absorption,
the breast consists mainly of two components: fibroglandular tissue and fat.
Fibroglandular tissue attenuates X-rays more than fat,
and,
as such,
is brighter on a radiograph of the breast,
while fat is darker.
Bright image regions are associated with fibroglandular tissue and the amount of “brightness” on breast images is referred as “breast density” (Figure 1).
The amount of fibroglandular tissue can be significantly different depending on the individual breast patterns,
and this can lead to different difficulties in detecting possible cancers,
mammography sensitivity decreasing while breast density increases.
Furthermore,
there is significant epidemiological evidence that extensive mammographic density is associated with an increased risk of developing breast cancer [1-3].
In Figures 2,
3,
and 4,
three different breast patterns are illustrated with increasing breast density.
Each sample case shows the four standard mammography views,
right and left cranio-caudal (CC) views and right and left medio-lateral oblique (MLO) views.
In clinical practice,
globally,
breast density is assessed by radiologists using various categorical scales (typically with 4-6 classes),
where the score increases with breast density [4].
The introduction of digital technologies in mammography allows measurement of quantitative breast density.
In particular,
Volpara is a commercially available software package which derives volumetric breast density automatically from a mammogram,
where the volumetric breast density is the percentage of fibro-glandular tissue in the breast; it is a true,
physical measure of breast composition.
Volpara works by starting from the individual pixel value and applying a physical model to describe the image formation process to estimate the x-ray attenuation between that pixel and the x-ray source.
From that attenuation,
you can work out how much fat and fibroglandular tissue was present [5-6].
Two examples of quantitative breast density maps associated respectively to a dense breast and a fatty breast are shown in Figures 5 and 6.
In these maps,
red is set to be high density and blue low density on an individual pixel basis.