Nashua, N.H.-based iCAD, Inc. has announced that the U.S. FDA has cleared ProFound AI, its deep-learning, cancer detection software for digital breast tomosynthesis (DBT), making the technology available for commercial sales and clinical use in the United States.

“Obtaining FDA clearance for ProFound AI opens a new and substantial addressable market for iCAD. This enables us to offer clinicians globally an unrivaled cancer detection and workflow solution built on the latest advances in deep-learning,” says Stacey Stevens, executive vice president and chief strategy and commercial officer at iCAD. “Clinical reader study results and comprehensive stand-alone testing have shown unprecedented improvements in both clinical performance and reading efficiency. We are proud to introduce revolutionary technology that will fundamentally transform breast cancer detection and patient care.”

The FDA clearance is based on positive clinical results from a large reader study completed earlier this year and presented at last month’s Radiological Society of North America annual meeting in Chicago. The research was performed with 24 radiologists who read 260 tomosynthesis cases both with and without iCAD’s ProFound AI solution. The findings show increased cancer detection rates, reduced false positive rates and patient recalls, and a decrease in interpretation times.

“This technology shows tremendous promise in assisting radiologists in detecting cancers, reducing recalls and increasing efficiency when reading tomosynthesis studies,” says Emily Conant, MD, professor and chief, division of breast imaging, vice chair of faculty development, department of radiology at the Hospital of the University of Pennsylvania. “Clinical data shows that when tomosynthesis readers use the ProFound AI algorithm, case-level sensitivity is improved by 8% on average and reading times are significantly decreased. Radiologists with various levels of expertise may benefit from this AI-driven technology when reading large tomosynthesis data sets.”

ProFound AI delivers critical benefits to radiologists, their facilities, and their patients through improvement of cancer detection rates by an average of 8% and decreasing unnecessary patient recall rates by an average of 7%, iCAD officials say. The new technology is trained to detect malignant soft-tissue densities and calcifications. It also provides radiologists with scoring information representing the likelihood that a detection or case is malignant based on the large dataset of clinical images used to train the algorithm.