Keywords:
Breast, Computer applications, CT, Mammography, Computer Applications-Detection, diagnosis, Screening, Segmentation
Authors:
J. Vandevenne1, C. Vanbilsen1, H. Bosmans2, S. De Buck2; 1Genk/BE, 2Leuven/BE
DOI:
10.26044/ecr2019/C-3683
Methods and materials
In this retrospective study concerning two age groups,
35 women (40-49 years) and 144 women (50-69 years) were included.
All patients underwent a CT thorax and a mammographic exam within a 12-month period.
Patients were asymptomatic concerning breast disease and had no history of breast cancer or biopsy.
Breast density was scored on mammograms according to BIRADS.
On CT Thorax,
breast density was measured with a treshold of 50 HU (two-compartment model separating glandular tissue and fatty tissue) after applying a semi-automated segmenting method (Syngo Via,
Siemens) (Figs.
1-3).
CT-BIRADS grades were defined as percentage of glandular tissue relative to total breast volume (grade 1: 0-25%; grade 2: 25.1-50%; grade 3: 50,1-75%; grade 4: 75,1-100%),
according to mammographic BIRADS grades.
Parametric correlation statistics (Pearson) were performed with a significance level of 0.05.