Nuance has combined its technology with NVIDIA's to better validate AI model development and deployment for medical imaging in clinical practice.
Nuance and NVIDIA join forces to accelerate clinical adoption of AI in imaging
November 22, 2022
by
John R. Fischer, Senior Reporter
To ensure that AI models are safe and effective for clinical practice in medical imaging, Nuance Communications is combining its radiology reporting platform with NVIDIA’s AI framework.
The two aim to provide, for the first time, AI-based diagnostic tools to radiologists and other clinicians at scale that can be refined to enhance patient care and lower costs.
The Nuance Precision Imaging Network is an AI-powered cloud platform, powered by Microsoft Azure, that provides access to AI-based tools and insights. MONAI is an open-source and domain-specialized framework for AI in medical imaging that was co-founded by NVIDIA.
The combined solution is expected to accelerate the adoption of AI applications, which has been hindered by complex workflows and a lack of standards, applications and deployment platforms.
"This partnership clears those barriers, enabling the extraordinary capabilities of AI to be delivered right at the point of care, faster than ever before,” said David Niewolny, director of healthcare business development of NVIDIA, in a statement.
NVIDIA’s radiology reporting system is used by more than 12,000 healthcare facilities for clinical workflows, as well as the 80% of U.S. radiologists that use its PowerScribe radiology reporting and PowerShare image sharing solutions.
MONAI is composed of MONAI Deploy, the accelerated processing pipeline that uses interoperability standards like DICOM, across data center and cloud environments, to easily integrate MONAI Applications Packages into healthcare systems.
Among the first to use the two together is Mass General Brigham, which has developed and deployed a breast density AI model that has reduced waiting time from several days to just 15 minutes. This decreases stress and anxiety associated with waiting.
It also has saved on costs associated with developing and maintaining AI applications for medical imaging, with continuous clinic-to-research feedback reducing model adaptation times from years to weeks and domain data shifts from months to weeks. Issue detection and repair also are capable in minutes rather than hours.
Dr. Keith Dreyer, chief data science officer of Mass General Brigham, says Nuance’s and NVIDIA’s technologies allow AI researchers to focus more on training and developing models. "That makes it simpler to get AI-powered insights to our clinicians, so they can provide the best possible care, accelerate time to treatment and improve patient outcomes."