Radboud university medical center accelerates medical imaging innovation with AWS Cloud

Radboud university medical center in partnership with Amazon (NASDAQ: AMZN) announce the launch of grand-challenge.org on AWS, a machine learning platform that accelerates innovation in the field of medical imaging by leveraging AWS Cloud solutions. The open-science platform allows builders from around the world to share sensitive medical images securely, collaborate globally on research projects, and build, benchmark, and deploy ML models to solve the hardest problems in the industry.

Despite billions spent globally every year on healthcare research and innovation, building AI/ML solutions for medical imaging remains a difficult and expensive undertaking. Researchers need access to high performance clusters of GPUs to run ML training jobs as well as large volumes of high-quality datasets, labeled and validated by experts. They also need to build, configure, and maintain complex AI/ML pipelines for training, tuning, debugging, and deployment.

Established by Radboud university medical center in 2010, grand-challenge.org allows researchers and clinicians to leverage AI/ML technologies, reducing costs, supporting collaboration, and accelerating innovation. With a sharp increase in demand for access to the platform during the COVID-19 pandemic, in March of 2020 Prof. Bram van Ginneken’s team at Radboud university medical center turned to AWS to migrate grand-challenge.org to the AWS cloud as well as expand the scope and scale of the platform. This partnership opened up new opportunities to accelerate the global medical imaging research community and today the platform has grown to over 64,000 registered users from across the world.

Since its inception, grand-challenge.org, has empowered medical teams and academic centers to reinvent their medical imaging research. Any research and development organization or consortium can organize medical imaging projects by publishing their research projects on grand-challenge.org. Participants from around the world gain access to curated data, tools to visualize and pre-process images, and end-to-end ML pipelines in one place, accessible from any location, through the convenience of a web browser. In addition, the global research community now has access to publicly available collection of AI/ML algorithms as well as medical image datasets made available at grand-challenge.org/algorithms/ and Registry of Open Data on AWS, respectively.

“Migrating grand-challenge.org to the AWS Cloud has allowed us to provide an integrated and scalable solution where medical researchers and data scientists can work together to build and validate AI/ML models,” said Prof. Bram van Ginneken of Radboud university medical center. “The AWS cloud can accelerate our research, and will allow us to bring potentially life-saving new image analysis algorithms to our healthcare systems.”

“We are delighted to be working with Radboud university medical center and grand-challenge.org to accelerate research in medical imaging,” said Razvan Ionasec, Technical Leader for Healthcare in Europe, Middle East, and Africa for Amazon Web Services. “grand-challenge.org on AWS will provide the technical and organizational framework required to allow builders and clinicians to innovate faster and collaborate more effectively to solve healthcare’s most difficult challenges.”

Since its launch, academic medical research centers have teamed up with grand-challenge.org to improve patient outcomes. Projects have included pathology diagnostics for prostate cancer, brain MRI lesion detection, lung cancer detection in chest radiographs and chest CT scans at institutions such as the Karolinska Institute, Amsterdam UMC, University College London, King’s College London, the National Institute of Health, Fraunhofer MEVIS, Vancouver General Hospital, and the Memorial Sloan Kettering Cancer Center.

To learn more visit: grand-challenge.org

• Want to know more about these subjects? Click on the buttons below for more news.

  News homepage news & mediaArtificial intelligence