How DeepHealth is Using AI to Screen for Breast Cancer

DeepHealth, a RadNet company, has introduced the DeepHealth Breast Suite – a modular AI platform built to strengthen breast cancer detection, risk stratification and clinical workflow efficiency.

Building on the company’s proven imaging technology, the Breast Suite integrates tools from iCAD to deliver a unified AI environment supporting more than 10 million mammograms annually.

“The launch of Breast Suite marks a pivotal step toward a new, AI-powered standard of care in breast cancer screening and diagnostic pathways," says Kees Wesdorp, President and CEO of DeepHealth.

“By embedding detection and risk intelligence with workflow tools, we give radiologists more capabilities to detect cancers earlier, with more confidence and to elevate patient care.”

Integrated AI for early detection

The DeepHealth Breast Suite combines AI-driven cancer detection, automated density assessment, risk modeling and in-development breast arterial calcification (BAC) analysis within one platform.

Its ProFound Pro detection engine enhances early disease identification through prior comparison data, automatic region-of-interest localisation and degree-of-suspicion scoring.

Additional features include automated density assessment for both 2D and 3D mammograms and a predictive, AI-based risk model estimating the likelihood of developing breast cancer within one to two years. According to DeepHealth, this model offers twice the accuracy of traditional questionnaire-based tools.

The company is also developing a new module to identify breast arterial calcifications, helping clinicians assess cardiovascular disease risk using existing mammogram data.

All features have undergone large-scale, real-world validation. A Nature Health study analysing mammograms from more than 579,000 women across over 100 community imaging sites found DeepHealth technology increased breast cancer detection by 21%, particularly among dense-breast and diverse populations.

The same study revealed that generalist radiologists achieved performance levels comparable to specialised breast imaging experts.

A separate Science Translational Medicine study evaluated DeepHealth’s risk assessment model in 154,000 European women. Researchers noted that offering additional screening to only the highest-risk 10% could have enabled earlier detection of 44% of cancers compared to 20% using the Tyrer-Cuzick model.

Workflow tools for modern imaging teams

In addition to detection and risk analysis, the Breast Suite integrates cloud-based workflow tools designed to help radiologists manage increasing imaging volumes.

Its cloud-first viewer supports multimodality imaging – including mammography, MRI, and ultrasound – and enables remote access.

Workflow features include prioritised worklists, rapid alerts for high-suspicion findings and the Safeguard Review workflow for secondary case review to reduce false negatives.

An intelligent reporting system supports guideline-based templates and automatically populates density findings for greater consistency.

Operating on DeepHealth’s AI-driven platform, the suite integrates seamlessly into healthcare environments while supporting secure remote access. The company confirms that continuous updates will ensure the system evolves with clinical needs.

Supporting risk-stratified screening

As imaging volumes rise and health systems move toward personalized screening strategies, the DeepHealth Breast Suite aligns AI-powered tools into a streamlined clinical ecosystem.

Now deployed across millions of annual mammograms, the suite is designed to enhance screening programmes by enabling more accurate, efficient and risk-based care decisions.

Its modular design offers flexibility for healthcare providers – allowing implementation of individual tools or deployment of the full suite as an integrated solution that adapts to diverse clinical and operational needs.