deepc and Vara Partner to Bring Proven AI for Breast Cancer Screening to Global Programs

AI in breast cancer screening is no longer judged by detection alone, but by its ability to perform reliably across entire populations over time.

deepc and Vara have entered a strategic partnership to expand access to AI developed specifically for breast cancer screening programs. As two German companies, they are combining clinical expertise and infrastructure to bring one of the most widely deployed AI solutions in this field to a broader international audience. Vara will make its technology globally accessible through the deepcOS® platform, extending a solution already embedded in Germany’s national screening infrastructure into broader clinical use, where it supports approximately 50% of the national program and has been validated in the PRAIM study, conducted across more than 460,000 women. The study demonstrated a 17.6% increase in cancer detection alongside a 15% reduction in recall rates, while reducing radiologist workload by more than half.

Breast cancer screening programs operate under strict requirements: large populations must be assessed consistently, often with double reading, across diverse sites and imaging systems. Most examinations are normal, yet each must be reviewed with the same level of care. Programs must balance two competing risks: missing cancers and overcalling healthy patients. At the same time, demand continues to grow while specialist capacity remains limited.

Delivering this level of consistent performance remains a broader industry challenge. These programs require AI solutions that can operate reliably across millions of cases under variable conditions, where differences in infrastructure, workflows, and patient populations are unavoidable. In this context, comparing findings across multiple prior screening rounds becomes critical, helping improve positive predictive value and reduce unnecessary recalls in longitudinal workflows.

Vara addresses these requirements with an AI solution built specifically for population-wide screening, where consistency, scale, and safety are critical. In Germany, its AI solution is widely deployed across the national breast cancer screening program and processes more than 200,000 cases per month. Over five years of continuous deployment and more than 6.1 million AI-supported reads reflect performance under routine clinical conditions across large-scale screening programs.

In programs where double reading is standard, Vara can act as an independent second reader for a substantial share of normal cases. This allows screening programs to maintain clinical quality while reducing reliance on scarce radiology capacity, addressing one of the core operational constraints in large-scale screening.

"For five years, we have focused on one thing: building the most reliable, most evidence-backed AI for breast cancer screening and proving it in the most demanding real-world environment possible. Today, approximately 50% of Germany's national screening program runs on Vara. With deepc, we now have a partner that shares our commitment to clinical rigor and understands what it takes to bring AI into radiology at scale. deepc's global platform and infrastructure give screening programs worldwide access to Vara's technology, and together, as two German AI companies born in radiology, we intend to set the standard for how AI supports breast cancer screening globally," said Jonas Muff, Founder and CEO of Vara.

Vara brings a strong foundation of clinical performance data and experience from large-scale, real-world deployment. Building on this, deepcOS® extends these capabilities into continuous, infrastructure-level monitoring across sites and workflows, enabling screening programs to track clinical performance metrics such as cancer detection and recall rates, compare performance over time, and maintain consistent standards across sites. This provides the level of visibility required to operate AI safely at a population scale.

Through this partnership, Vara's screening AI solution can now be deployed more broadly across healthcare networks. deepcOS® provides the vendor-neutral infrastructure to integrate AI solutions into existing clinical and IT environments, support consistent deployment across sites, and govern performance continuously once solutions are live. This enables screening programs to adopt a solution already validated at a national scale, while maintaining operational control and visibility across their own networks.