Google Health’s mammography AI models will be integrated into real-world clinical practices for the first time through a commercial licensing agreement with a global leader in 3D mammogram screening technology, iCAD, the companies announced on Monday.
Breast cancer screening access, accuracy, patient experience, and clinician workload could be improved by the partnership.
Breast cancer remains challenging to detect and diagnose despite the widespread use of digital mammography. On average, over the course of ten years of mammograms, there is a 50-60 percent chance of getting at least one false positive result. In many cases, double-readings for mammography is difficult due to a global shortage of mammography professionals.
With the help of clinical research partners, Google began developing and testing AI models that could improve breast cancer detection in 2018.
Google’s AI technology showed early success, with a report in Nature Magazine in 2020 stating that AI can detect breast cancer signs better than radiologists, while also reducing false positives.
Additionally, iCAD has developed its own AI-powered screening technology, Breast AI, which is also showing early promise.
Through the partnership, iCAD would leverage Google’s AI and cloud technology to enhance iCAD’s existing systems. For instance, there may not be enough computing power in some regions to run AI systems. The problem can be solved by using cloud-based infrastructure.
“iCAD will work toward validating and incorporating our mammography AI technology with its products for use in clinical practices with the goal of improving breast cancer detection and assessment of short-term personal cancer risk for the more than two million people globally diagnosed with breast cancer every year,” Greg Corrado, Head of Google’s Health AI initiatives wrote in a statement.