Digital Mammography versus Breast Tomosynthesis: Impact of Breast Density on Diagnostic Performance in Population-based Screening

Abstract

he sensitivity of digital mammography (DM) is lower in women with dense breasts than in those with lower breast density (1). Breast density is also associated with higher false-positive rates and recall rates (2) due to superposition of normal glandular tissue that can mimic cancer. The woman's age has an impact on mammography screening as breast density decreases (3) and cancer incidence increases. The distribution of cancers shifts toward less-aggressive slower-growing cancers with increasing age (4). It has been shown that mammography screening has a lower sensitivity (1) and higher false-positive rate (2) among younger women. Digital breast tomosynthesis (DBT) generates pseudo three-dimensional (3D) images where a single section of anatomy is in focus. The rest is blurred, with greater magnitude proportional to the distance from the focus plane. The screening performance of DBT for specific density and age groups may be different from that of DM, as DBT potentially can reduce masking and resolve superposition of breast tissue. Prospective (5-11) and retrospective (12-18) studies have shown that the integration of DBT improves the cancer detection or recall rates for both fatty and dense breasts and in age groups relevant for mammography screening. Data are limited in almost entirely fatty and extremely dense breasts. Two large studies compared DBT and DM in women with extremely dense breasts, with one study finding an increased cancer detection rate with DBT (5) and the other finding similar rates for DBT and DM (13). Therefore, there is a need for more data from large prospective trials.

Average glandular dose in paired digital mammography and digital breast tomosynthesis acquisitions in a population based screening program: effects of measuring breast density, air kerma and beam quality

“…Therefore, they will represent women with cancers, benign findings or women that are difficult to assess. This group has a higher breast density on average than all women included in OTST (155).…”

“…This indicates that the radiologists were more likely to classify breasts as dense and extremely dense in the reader study than in a screening setting. (155). As the confidence intervals overlap, the difference was not significant.…”

“…A thorough description of the study design and image interpretation has been reported elsewhere (41,108,154,155). All women with at least one positive interpretation were discussed at a consensus meeting, where the radiologists evaluated breast density using BI-RADS density 4 th edition in consensus.…”

Stand-alone artificial intelligence - The future of breast cancer screening?

“…Specifically, let’s consider the scenario of replacing digital mammography as the screening imaging modality for a higher-performing technique, e.g., digital breast tomosynthesis (DBT). Over a number of prospective screening trials, DBT has been found to result in higher detection performance, but with a doubling of the reading time [ [21] , [22] , [23] , [24] , [25] , [26] , [27] , [28] , [29] ]. Therefore, introduction of DBT-based screening, which has already taken place in the US, where screening is institution-based [ 30 ], is also of intense interest in countries where screening is regional- or national-program based.…”

Mass Detection and Segmentation in Digital Breast Tomosynthesis Using 3D-Mask Region-Based Convolutional Neural Network: A Comparative Analysis

“…For example, the DBT increases the cancer detection rate but is less effective for women with extremely dense breasts (Vourtsis and Berg, 2019). A recent study reported that DBT enabled the detection of more cancers in all density and age groups compared with DM, especially cancers classified as spiculated masses and architectural distortions (Osteras et al, 2019). DBT and DM screening increased the detection rate of histologically favorable tumors compared with that attained by DM screening (Hofvind et al, 2018).…”