NVIDIA, often considered one of the major companies helping to drive the rapid development of AI technology through its line of GPU-based hardware and associated software products, is pushing heavily into the field of AI-assisted radiology.
The company is expanding its Clara Platform which allows users to utilize virtual super computing capabilities to advance medical imaging and has signed on new strategic partners to accelerate the adoption of AI in medical imaging.
NVIDIA has rolled out its Clara Software Development Kit (SDK), which allows third-party developers to build advanced imaging applications on top of the Clara Platform. The SDK contains a set of accelerated libraries, engines that utilize the libraries to preform higher level functions that developers can leverage and the ability to provide remote access to visualizations for patients.
“The idea is that you build it once and can deploy it everywhere on your imaging chain,” said Abdul Hamid Halabi, NVIDIA’s global business development lead for healthcare and AI, in a call with reporters.
The company says its working with 75 partners in the healthcare space including medical centers, healthcare startups, research centers and medical imaging companies.
In order to boost the efforts of developers to build and use these AI applications, NVIDIA has also announced its Transfer Learning Toolkit, which allows physicians to customize AI applications and models for their own practice by incorporating a segment of their private patient data into the larger algorithm through a process of AI-assisted annotation. The software feature is expected to be available in early 2019.
A Deep-dive Into Specialty Pharma
A specialty drug is a class of prescription medications used to treat complex, chronic or rare medical conditions. Although this classification was originally intended to define the treatment of rare, also termed “orphan” diseases, affecting fewer than 200,000 people in the US, more recently, specialty drugs have emerged as the cornerstone of treatment for chronic and complex diseases such as cancer, autoimmune conditions, diabetes, hepatitis C, and HIV/AIDS.
Halabi said the toolkit was informed by the company’s initiatives in autonomous vehicles which involve localizing algorithms based on the market in which they are deployed.
“Every radiology practice has its own instruments, has its own patients, demographics and its own way of practice. So although we’re seeing a flood of new algorithms come through, we do believe we need to provide radiologists with the tools to take those AIs and localize them for their own patients,” Halabi said. “It’s going to help most radiologists unlock the value of the data that they’re sitting on today.”
Alongside their expanded offerings in radiology software, the company has also inked two deals intended to accelerate the adoption of AI technology in the field.
One is an effort with The Ohio State University’s Wexner Medical Center to build what NVIDIA is calling the first “in-house” marketplace for clinical medical imaging in the U.S.
Essentially, the company has helped OSU build a local app store to host various imaging algorithms which can be quickly integrated across the hospital for various application including the detection of a brain hemorrhage or coronary artery disease.
The idea is that the technology can be integrated across a variety clinical workflows ranging from an early warning system for the ER department to a diagnostic aid in the imaging reading room.
NVIDIA is also broaching the clinical trial space through a partnership with the National Institutes of Health, one of the world’s leading centers for biomedical research.
The first stage of the partnership will involve NVIDIA researchers working with clinicians at the NIH Clinical Center in Bethesda, Maryland on projects involving the use of AI imaging tools to streamline clinical trials in oncology.
The research efforts will be focused on speeding up the research process for brain and liver cancer treatments and developing tools to combine imaging, genomic and clinical data for precision medicine efforts.
A larger goal of the partnership is to improve on the largely manual process of cancer staging by using AI to more accurately characterize and measure tumors and incorporate individual bio marker data.
“We all know that today’s cancer staging is done based on standards that were set by physicians a long time ago and estimated measurements of the tumors,” Halabi said. “We believe that with AI we can get closer to precision health by working with NIH to bring in all the different data available to us – from genomics to imaging – for better cancer staging.”
Picture: Getty Images, wigglestick
