This poster is published under an
open license. Please read the
disclaimer for further details.
Keywords:
Breast, Mammography, Observer performance, Comparative studies, Cancer, Epidemiology, Image registration
Authors:
R. R. Winkel1, M. von Euler-Chelpin1, M. Nielsen2, M. Bachmann Nielsen1, W. Uldall1, P. Diao1, I. Vejborg1; 1Copenhagen/DK, 2Copenhagen /DK
DOI:
10.1594/ecr2015/C-1037
Aims and objectives
Breast density or paranchymal pattern of the breast are major risk factors for breast cancer with a four to six-fold increased risk of breast cancer for high-density women [1],
[2].
This increased risk so-exists with a reduction in mammographic sensitivity—the so-called masking effect—besides density being an independent marker of risk.
In recent years personalized screening strategies based on a woman´s risk profile (density profile) have been in focus.
However,
a great number of methodologies for measuring density or parenchymal pattern on mammograms exist.
These methodologies include different visual classifications [3]–[5] as well as several newer semi or fully-automated techniques [6]–[8].
The manual (visual and semi-automated) breast density methods do not require very advanced or expensive technical equipment but are greatly influenced by subjectivity.
We aimed to report inter-observer agreement on three different manual ways of assessing mammographic density and parenchymal pattern,
and to examine what impact inter-observer variability has on breast cancer risk prediction in terms of odds ratios.