Riverain Technologies Advances Clinical A.I. Through Use of Synthetic Training Data
ClearRead CT is an FDA-Cleared, Patent-Protected, Deep Learning Radiology Solution
Dayton, Ohio – July 17th, 2018 – ClearRead CTTM is the only FDA-cleared (September 2016) radiology application developed using synthetic data for training a deep learning solution. The concurrent read device aids clinicians in the detection of solid, part solid and ground glass nodules – all early indicators of lung cancer. A pivotal blind multi-reader, multi-case reader study demonstrated that readers using ClearRead CT read 26% faster and detected 29% of previously missed nodules1. Key to achieving the performance was the novel, patent-protected ClearRead CT vessel suppression technology that transforms a standard chest CT into a vessel free series. Results related to the application have been published in the American Journal of Roentgenology1 (March 2018, Volume 210) and in the European Journal of Radiology2 (April 2018, Volume 101).
A fundamental challenge in using artificial intelligence in radiology is the availability of imaging data sets of a scale and precision necessary to optimize machine learning solutions. Riverain has addressed this challenge through the development of synthetic data modeling tools that allow massive training sets to be developed with the necessary supporting data. These advances have led to broad clinical acceptance and accelerating commercial implementation of the company’s software. Revenue in 2018 has grown 94% year-over-year as the company’s tools are increasingly used by radiologists globally in leading health systems such as Duke University, Einstein Medical Center, University of Michigan, University of Maryland, University of Chicago and top Veterans Affairs medical centers.
“Using Riverain’s unique methods, the ClearRead CT software provides proven results, instilling confidence in the clinical decision process,” said Riverain Technologies CEO, Steve Worrell. “Our goal is global widespread adoption and we are well on our way to making a major impact in the Clinical AITM space.”
1. American Journal of Roentgenology: 1-8. 10.2214/AJR.17.18718
2. Vessel suppressed chest Computed Tomography for semi-automated volumetric measurements of solid pulmonary nodules. Milanese, Gianluca et al. European Journal of Radiology, Volume 101, 97 – 102.