Tuesday, November 27 | 3:30 p.m.-3:40 p.m. | SSJ01-04 | Arie Crown Theater
What does automated breast ultrasound (ABUS) contribute to cancer screening when it's used with mammography in women with dense breasts? Researchers from the University of Chicago will present findings that address this question on Tuesday afternoon.Maryellen Giger, PhD, and colleagues gathered a cancer-enriched set of screening mammography and ABUS exams from asymptomatic women with breast density BI-RADS 3 or 4. Seventeen radiologists first interpreted the mammography images alone, then interpreted the mammography images with ABUS.
Giger's team used area under the receiver operator characteristics curve (AUC) to assess readers' performance with mammography alone versus mammography plus ABUS.
Mean AUC across all readers was 0.604 for mammography alone and 0.747 for mammography plus ABUS, the group found. Specificity was comparable, at 78% for mammography alone and 76% for mammography with ABUS. But ultrasound improved the overall sensitivity across all readers, from 38.8% for mammography alone to 63.1% for mammography plus ABUS, an increase of 23.3 percentage points.
Does automated breast ultrasound boost mammography's performance? Yes, Giger's group concluded, and the technology can be expected to benefit women with dense breast tissue by providing earlier detection of cancers missed at mammography.










![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=112&q=70&w=112)






