March 27, 2017 — iCAD announced that PowerLook Tomo Detection received Premarket Approval (PMA) from the U.S. Food and Drug Administration (FDA). PowerLook Tomo Detection is a concurrent-read computer-aided detection (CAD) solution for digital breast 3-D tomosynthesis, available on the PowerLook Breast Health Solutions platform.
Two-dimensional digital mammography typically produces four images per exam while digital breast 3-D tomosynthesis can produce hundreds of images, significantly increasing exam interpretation time for radiologists. PowerLook Tomo Detection improves radiologists’ efficiency by automatically analyzing each tomosynthesis plane and identifying suspicious areas. The suspicious areas are naturally blended onto a 2-D synthetic image to provide radiologists with a single enhanced image that is used to more efficiently navigate the large tomosynthesis data set.
“iCAD has taken a refreshing new approach to computer-aided detection. This innovative workflow solution detects suspicious areas on the tomosynthesis planes and that information is used to deliver an enhanced image that focuses the radiologist on the specific areas that need further investigation,” said Justin Boatsman, M.D., medical director and diagnostic radiologist, Intrinsic Imaging LLC, who took part in the U.S. clinical study. “This not only helps reduce the reading time and improve the reading experience for radiologists, but it can also provide radiologists with an added level of confidence.”
In a U.S., clinical study conducted from October 2015 to January 2016, radiologists were able to significantly reduce reading time when reading 3-D tomosynthesis exams with PowerLook Tomo Detection. The study included 20 radiologists reading 240 tomosynthesis cases, both with and without the PowerLook Tomo Detection technology. Reading time was reduced by up to 37 percent with an average reduction of 29 percent when using PowerLook Tomo Detection, with no statistically significant impact on sensitivity, specificity or recall rate.
Another European clinical study was completed with six radiologists reading 80 cases and showed similar results and was the basis for CE Mark of PowerLook Tomo Detection in April 2016.
The application employs deep learning, a branch of machine learning that uses sophisticated algorithms that are trained to recognize the visual characteristics of a cancer by analyzing actual patient images. The current version of the algorithm was trained using thousands of images.
PowerLook Tomo Detection, currently available on GE Healthcare digital breast tomosynthesis systems, also received CE Mark and Health Canada approval in 2016, and is currently being used by multiple high volume breast imaging centers in Europe.
For more information: www.icadmed.com