Sunday, December 1 | 11:20 a.m.-11:30 a.m. | S2-SSBR01-6 | Room S406A
Here, results will be shown on how AI impacts breast cancer detection on digital breast tomosynthesis (DBT) exams based on various factors.
In her talk, Kathy Colletta, MD, from Boca Raton Regional Hospital in Florida will present her team’s findings, which demonstrate that AI boosts overall cancer detection rates, invasive and lobular detection rates, and cancers detected in dense breasts.
The research team performed a retrospective analysis of mammography audit data and screening cancers detected at four sites during two time periods with nine breast radiologists. The first time period spanned from 2018 to 2020, before the radiologists had access to AI. The second time period spanned from 2020 to 2022, during which the radiologists used AI (ProFound Detection, iCAD).
The team collected data from all screen-detected cancers regarding age, breast density, tumor size, staging, and histopathology.
The pre-AI period had 54,440 exams (339 true positives) and the post-AI period had 48,742 exams (369 true positives). The cancer detection rate per 1,000 improved from 6.23 to 7.57 between the two periods (p < 0.01). Also, AI led to a 7.8% increase in cancers detected in dense breasts (p = 0.04).
Among other findings, the average size of invasive cancers found in the post-AI period decreased by nearly 1.5 mm compared to those found during the pre-AI period. Also, more invasive cancers were T1 post-AI than pre-AI, an increase of 7% (p = 0.01), without any change in the detection of ductal carcinoma in situ (DCIS) post-AI.
What else did the team find that points to AI’s utility in this area? Attend this presentation to find out.
