Visage 7 Is Prominently Positioned on the AI Pipeline, Key to Driving the Visage AI Accelerator Solution
Visage Imaging, Inc., a wholly owned subsidiary of Pro Medicus Ltd., has announced Visage will demonstrate semantic annotations for the Visage 7 Enterprise Imaging Platform as a works-in-progress, at the Radiological Society of North America (RSNA) 2019 annual meeting (Visage Booth 7524, McCormick Place-North Hall), December 1-5, in Chicago, Illinois.
In the age of artificial intelligence (AI), there is tremendous and increasing value in the data that radiologists produce as part of their reading workflow where they interpret and annotate imaging studies. This value can only be fully realized if the resulting data is mineable and its semantic meaning is computer accessible.
“That is why Visage is developing a framework to make annotations semantically meaningful,” explained Malte Westerhoff, PhD, Visage Imaging Co-Founder and Chief Technology Officer. “Then these annotations, generated during routine diagnostic interpretation, can flow into a model training infrastructure to build new AI models. The algorithms output from those models can then be integrated into the radiologist’s diagnostic viewer.”
Visage is working with the American College of Radiology (ACR) Data Science InstituteTM to integrate Visage 7 semantic annotations with the ACR AI-LABTM, which offers radiologists tools designed to help them learn the basics of AI and participate directly in the creation, validation and use of health care AI.
Dr. Westerhoff continued, “Visage 7 is prominently positioned on both sides of the AI pipeline. Visage is developing, together with our partners, the framework and user interface to seamlessly create semantic annotations directly in Visage 7, to exchange those annotations with other third-party applications, including the ACR AI-LAB, and finally, to consume the output of AI models and integrate them directly into Visage 7.”
A pilot implementation of Visage 7 semantic annotations is ongoing at Partners HealthCare Massachusetts General Hospital (MGH) and Brigham and Women’s Hospital (BWH) Center for Clinical Data Science, where the framework is being defined and tested in the radiology workflow. “This is an exciting effort that allows radiologists to provide highly meaningful contributions for model development without interrupting their workflow,” said Keith Dreyer, DO, PhD, Chief Data Science Officer, Partners HealthCare.
Visage 7 semantic annotations is the newest component of the Visage AI Accelerator, a new end-to-end solution from Visage that dramatically reduces the time from research to bedside. Visage AI Accelerator includes joint development and commercialization opportunities with academic customers and third-parties; the Visage 7 AI research server that streamlines AI algorithm development with an open API and unifies research and diagnostic imaging in a singular platform (works-in-progress); and an AI ecosystem of third-party and Visage-developed algorithms.