A new method for quantitative analysis of mammographic density.

摘要: Women with mammographic percent density >50% have a approximately three-fold increased risk of developing breast cancer, potentially making them screening candidates for MRI scanning. The purpose this work is to introduce new method quantify (MPD), and compare the results current standard care assessment. Craniocaudal (CC) mediolateral oblique (MLO) mammograms 104 patients were digitized analyzed using an interactive computer-assisted segmentation routine implemented two purposes: (1) segment area from background radiographic markers, (2) dense fatty portions breast. Our technique was evaluated by comparing qualitative estimates determined certified radiologist BI-RADS Categorical Assessment (1 (fatty) 4 (dense) scale). Statistically significant correlations (two-tailed, p < 0.01) observed between calculated MPD both CC (Spearman rho = 0.67) MLO views 0.71). For view, statistically differences revealed mean each category except (BI-RADS 1) scattered 2). Finally, views, in all categories observed. Comparing strong positive correlation (Pearson r 0.8) MPD. In addition, evaluation reproducibility our demonstrated average deviation subsample eight patients, measured five times, 1.9% (range: 0.03%-9.9%). Eliminating one misassignment reduced 0.75% 0.03%-3.16%). Further analysis 10% patient sample agreement (ICC 0.80-0.85) reliability views. Overall, these demonstrate feasibility utilizing approach quantitative segmentation, which may be useful detecting small changes introduced through chemoprevention, diet, or other interventions.