Mammographic breast density distribution and the fifth-edition BI-RADS criteria

In the present study, we measured breast density by using a fully automated software and cross-compared density distribution functions of BI-RADS categories. Our final goal was to understand to what extent the new BI-RADS categories represent mammographic density distribution. The density distributions of the BI-RADS categories measured with a publicly available software tool show that: category A dominates up to 11% with the probability P similar to 0.90 to have such density; category B dominates in the interval between 11 and 17% with P similar to 0.25 to have a density within this region; category C predominate between 17 and 32% with the probability P similar to 0.53 to find breast inside this interval; category D is dominant for densities higher than 32% with P similar to 0.86 to have such density.