RadNet, Hologic Collaborate on AI for Breast Health

RadNet Inc., a national leader in providing high-quality, cost-effective, fixed-site outpatient diagnostic imaging services, and Hologic, Inc., an innovative medical technology company primarily focused on improving women’s health, have entered into a definitive collaboration to advance the use of artificial intelligence (A.I.) in breast health.

As the world leader in mammography, Hologic will contribute capabilities and insights behind its market-leading hardware and software, and will benefit from access to data produced by RadNet’s fleet of high-resolution mammography systems, the largest in the nation, to train and refine current and future products based on A.I. RadNet will share data from its extensive network of imaging centers, as well as provide in-depth knowledge of the patient pathway and workflow needs to help make a positive impact across the breast care continuum. The collaboration will enable new joint market opportunities and further efforts to build clinician confidence and develop and integrate new A.I. technologies.

“We believe the future of breast health will rely heavily on the integration of A.I. tools, such as our 3DQuorum imaging technology, as well as next generation CAD software, that aid in the early detection of breast cancer,” said Pete Valenti, Hologic’s division president, Breast and Skeletal Health Solutions. “We are energized by the opportunities this transformative collaboration with RadNet creates for patients and clinicians alike. Access to data is critical in training and refining A.I. algorithms. With this collaboration, we now have the opportunity to leverage data from the largest fleet of high-resolution mammography systems to develop new tools across the continuum of care, provide workflow efficiencies, and improve patient satisfaction and outcomes.”

As part of its collaboration with Hologic, RadNet intends to upgrade its entire fleet of Hologic mammography systems to feature Hologic’s 3DQuorumÔ imaging technology, powered by Genius AIÔ. This technology works in tandem with Clarity HDÔ high resolution imaging technology to reduce tomosynthesis image volume for radiologists by 66 percent.¹ Additionally, all of RadNet’s Hologic systems are anticipated to feature the Genius3D Mammography exam, the only mammogram clinically proven and FDA approved as superior for all women, including those with dense breasts, compared with 2D mammography alone.^2-5

The collaboration will be bolstered by RadNet’s recent acquisition of DeepHealth (Cambridge, MA), which uses machine learning to develop software tools to improve cancer detection and provide clinical decision support. Led by Dr. Gregory Sorensen, DeepHealth’s team of A.I. experts is focused on enabling industry-leading care by providing products that clinicians and patients can trust. In addition, the DeepHealth team will integrate its A.I. tools within the Hologic ecosystem. “When seeking a partner and reviewing options amongst all mammography vendors, we selected to integrate our tools with Hologic’s market-leading technology,” said Dr. Sorensen. “Hologic’s systems produce the highest level of spatial resolution in the market. Hologic also has the largest domestic footprint and market share in 3D Mammography systems. This integration will allow the DeepHealth team to train its algorithms for use with the most advanced screening technology possible. As Hologic and RadNet share their respective capabilities and tools, greater efficiency and accuracy can be achieved by our radiologists.”

“Much like RadNet, Hologic is a highly innovative company and market leader in breast health,” said Howard Berger, MD, RadNet’s chairman and CEO. “When Hologic’s leading screening technology is paired with RadNet’s approximately 1.2 million annual screening mammograms, the resulting dataset becomes a powerful tool to train algorithms. We see the future as being transformative for both of our organizations.”

“We have witnessed how the application of our Genius AI technology platform has improved cancer detection, operational efficiency and clinical decision support across the breast cancer care continuum,” said Samir Parikh, Hologic’s global vice president for research and development, Breast and Skeletal Health Solutions. “We look forward to building upon these advances in collaboration with Dr. Sorensen and the RadNet team to expand the use of machine learning, big data applications and automated algorithms impacting global breast care.”

References
¹ Report: CSR-00116
² Results from Friedewald, SM, et al. “Breast cancer screening using tomosynthesis in combination with digital mammography.” JAMA 311.24 (2014): 2499-2507; a multi-site (13), non-randomized, historical control study of 454,000 screening mammograms investigating the initial impact the introduction of the Hologic Selenia Dimensions on screening outcomes. Individual results may vary. The study found an average 41% increase and that 1.2 (95% CI: 0.8-1.6) additional invasive breast cancers per 1000 screening exams were found in women receiving combined 2D FFDM and 3D™ mammograms acquired with the Hologic 3D Mammography System versus women receiving 2D FFDM mammograms only.
³ Freidewald SM, Rafferty EA, Rose SL, Durand MA, Plecha DM, Greenberg JS, Hayes MK, Copit DS, Carlson KL, Cink TM, Carke LD, Greer LN, Miller DP, Conant EF, Breast Cancer Screening Using Tomosynthesis in Combination with Digital Mammography, JAMA June 25, 2014.
⁴ Bernardi D, Macaskill P, Pellegrini M, et al. Breast cancer screening with tomosynthesis (3D mammography) with acquired or synthetic 2D mammography compared with 2D mammography alone (STORM-2): a population-based prospective study. Lancet Oncol. 2016 Aug;17(8):1105-13.
⁵ FDA submissions P080003, P080003/S001, P080003/S004, P080003/S005