by John R. Fischer
, Senior Reporter | October 04, 2018
CureMetrix has agreed to provide its AI-based computer aided detection (AI-CAD) software for mammography to DocPanel, an online community of expert subspecialty radiologists, as part of a new collaboration.
The two aim to use the CureMetrix cmAssist solution to help DocPanel radiologists produce second opinions of mammograms and provide more accurate diagnoses for breast cancer patients throughout the U.S. and abroad, sparing them the costs and time of traveling to experts for diagnoses.
“As the premier AI solution in the world for breast imaging, cmAssist was implemented to study the impact of a high-quality CAD that works in teleradiology,” Kevin Harris, CEO of CureMetrix, told HCB News. “DocPanel's focus is to deliver the best care possible to their patients and imaging practices and our focus is to enable them with the highest-quality technology to do so.”
An investigational solution, CureMetrix cmAssist utilizes AI
to help radiologists identify suspicious areas of interest on mammograms for earlier and more accurate detection of breast cancer. The algorithms behind its functioning have been proved to detect suspicious lesions six years before they were confirmed to be cancer in biopsies. The solution is also equipped with neuScore, a feature that helps clinicians understand the level of suspicion.
Utilizing DocPanel’s online platform, patients gain access to hard-to-reach expert breast radiologists in subspecialty niches who can use CureMetrix’s AI-CAD technology to make examinations and form second opinions on the condition of a patient from anywhere in the world.
The partnership is not the first for CureMetrix this year, having recently teamed up with the University of Florida for the development of next-generation CAD tools
for mammographers. It also provided the university with access to its physics-based AI and deep-learning software solutions as part of an educational and investigational endeavor that aims to evaluate its efficiency and impact in producing accurate readings of breast imagery.
In addition to detecting suspicious lesions in advance, CureMetrix cmAssist has produced fewer false positives than current CAD products currently on the market, according to Harris.
“cmAssist has fewer false positives due to the algorithm being trained and validated on over 500,000 images from diverse subsets of the population and a heavy reliance on our strong cohort of medical advisers to better understand what the doctors need to see, which characteristics pose the most challenges and what true positives look like,” he said.
DocPanel is equipped with more than 300 board-certified and fellowship-trained radiologists spread out across 41 states and who are trained to cover every modality and specialty in radiology. It is available on a global scale, with approximately 30 percent of second opinion cases uploaded by international patients.