Transpara Breast AI at RSNA 2025: Evidence, efficiency, and practical gains for breast imaging programs
ScreenPoint Medical is presenting the Transpara Breast AI suite at the 111th RSNA Annual Meeting (Nov 30-Dec 4, South Hall #4719). The suite brings Detection, Density, and Temporal Comparison into one workflow-ready platform that serves as a second set of eyes for radiologists. The focus is simple: detect cancers earlier, reduce reading burden, and improve screening program performance at scale.
Why this matters for your program
2025 has been a strong validation year for Transpara in both research and routine use. The platform has processed over 11 million mammograms and is deployed across 30+ countries. In the U.S., 40% of the Top 20 Best Hospitals Honor Roll recipients (US News & World Report, 2025) have selected Transpara.
Critically, Transpara is part of the only randomized controlled trials in Breast AI: the MASAI RCT in Sweden and the upcoming $16M PRISM RCT in the United States led by UCLA and UC Davis (starts early 2026). MASAI results published earlier this year reported a 29% increase in cancer detection and a 44% decrease in screen-reading workload with Transpara Detection compared to standard of care.
Risk assessment takes center stage
At RSNA, Transpara Risk-a 5-year, image-based breast cancer risk model-and Transpara Detection anchor new evidence: "Added Value of Breast Cancer Risk Prediction Versus Detection Over a Two-year Time Period with Mammography" (S5-SSBR02-4, Sunday, Nov 30, 2:30 PM, S406A). Findings show added value for precision care, with Transpara solutions outperforming both RSNA CAD and Mirai in this research. Transpara Risk is approved for investigational use only.
Where to see the data at RSNA 2025
- Comparative performance on DBT
"Comparing Four Commercial AI Algorithms for Standalone Breast Cancer Detection on Digital Breast Tomosynthesis in a Dutch Population-based Screening Cohort" (T3-SSBR05-6, Tuesday, Dec 2, 9:30 AM, S406A). First head-to-head comparison on the same large dataset; Transpara Detection outperformed other commercial algorithms. - Interval cancers
"Potential of Artificial Intelligence to Detect Interval Breast Cancers" (M5A-SPBR, Monday, Dec 1, 12:15 PM, Learning Center). Suggests Transpara Detection may have supported earlier detection of 38% of interval cancers, especially those missed or underestimated by radiologists. Research conducted at Johns Hopkins Medicine. - Specificity gains using prior exams
"Use of Multiple Prior Exams Improves Specificity of an AI System for Breast Cancer Detection in a Retrospective Multi-site Validation" (T3-SSBR05, Tuesday, Dec 2, 9:30 AM, S406A). Higher specificity when prior exams are incorporated into Transpara analysis. - Real-world adoption
"Real-World Performance of an FDA-approved Artificial Intelligence Tool for Screening Mammography" (W5B-SPBR-3, Wednesday, Dec 3, 12:45 PM, Learning Center). Early implementation in a busy screening practice showed strong risk stratification and support for recall decisions. Research conducted at Johns Hopkins Medicine. - DBT pre-reading to reduce workload
"AI Pre-reading of Digital Breast Tomosynthesis: A Multi-site Validation of a Novel Imaging Biomarker for Confident Identification of Normal DBT Examinations" (R1-SSBR10, Thursday, Dec 4, 8:00 AM, S406A). A novel AI-driven biomarker identified ~one-third of normal DBT exams with a negative predictive value near 100%, enabling safe pre-reading and faster throughput.
Practical takeaways for clinical leaders
- Screening efficiency: Use AI pre-reading and risk stratification to route low-risk, normal-appearing DBT exams faster while preserving safety.
- Reader workload: Evidence points to meaningful reductions in screen-reading time without sacrificing detection, especially in high-volume programs.
- Quality and specificity: Incorporate multiple priors into AI analysis to improve specificity and reduce unnecessary recalls.
- Program performance: Combine Detection, Density, and Temporal Comparison to reinforce consistency across readers and sites.
- Risk-guided pathways: Image-based 5-year risk can enable personalized screening intervals and targeted prevention strategies (Transpara Risk is investigational).
- Integration: Deploy within enterprise workflows (PACS/VNA/RIS) to ensure AI outputs appear where radiologists already work.
Visit ScreenPoint Medical at RSNA
See Transpara Breast AI in action at Booth #4719 (South Hall). For details on sessions and logistics, visit the RSNA Annual Meeting site. To schedule an online product overview after the show, visit screenpoint-medical.com.
About ScreenPoint Medical
ScreenPoint builds AI-driven technology for the breast imaging continuum-supporting consistent reading, reducing uncertainty, and improving patient experience. Proven through research and real-world use, Transpara Breast AI is trusted by radiologists and programs worldwide.
Your membership also unlocks:
