AI-Driven Breast Cancer Screening Boosts Detection by 21.6% in Major US Study

November 17, 2025
AI-Driven Breast Cancer Screening Boosts Detection by 21.6% in Major US Study
  • RadNet leadership emphasizes scale, diversity, and real-world relevance, portraying the program as delivering equitable results across racial, ethnic, and density groups.

  • Authors and affiliations indicate collaboration between DeepHealth and academic partners, with disclosures of corporate involvement and potential conflicts of interest.

  • Dr. Gregory Sorensen highlights the real-world, community-center setting and diverse population as strengths, noting AI improves access to specialist-level care.

  • RadNet media contact information is provided for further inquiries.

  • Looking ahead, the program plans continued outcome reporting and site-level performance summaries at RSNA 2025, with near-term metrics on site adoption and aggregated detection/recall data and longer-term signals on stage shift or survival benefits.

  • Data underpinning the findings are controlled by a supervising institutional review board and are not publicly accessible, as they are part of DeepHealth’s intellectual property.

  • The article underscores potential implications for broader access to specialist-level care and the benefits of early cancer detection regardless of location.

  • The ASSURE study, the largest real-world US analysis of AI-driven breast cancer screening, found RadNet and DeepHealth’s AI workflow increased cancer detection by 21.6% compared with state-of-the-art 3D mammography, raised positive predictive value by 15%, and kept recall rates within ACR guidelines, with benefits consistent across populations including over 150,000 Black women and a 22.7% boost for dense breasts.

  • EBCD™ is powered by DeepHealth’s Breast Suite, and the program has rolled out nationwide at RadNet-affiliated centers since 2023 to detect suspicious lesions earlier and improve treatment options.

  • The AI workflow combines an AI-based detection/diagnosis tool with a safeguard review where at-risk cases receive additional radiologist review to enhance early cancer detection.

  • The release includes forward-looking statements with cautionary language about potential risks and regulatory considerations that could affect realized benefits.

  • Standard forward-looking risk disclosures warn of regulatory, economic, and technological conditions that may influence outcomes.

Summary based on 8 sources


Get a daily email with more AI stories

More Stories