Exciting news! Our Co-founder and Chief Operating Officer, Chenyu Wang, was recently invited to speak at NVIDIA’s GTC23 about the latest advances in federated learning for medical imaging applications.
23 March 2023
On March 23rd, Dr. Chenyu Wang, co-founder and Chief Operating Officer of SNAC, joined other experts at the GTC23 technology conference hosted by NVIDIA to discuss the latest developments and clinical applications of federated learning in their respective fields. Along with Peter Grandsard, Executive Director of R&D at Amgen, and Ittai Dayan MD MPH, CEO of Rhino Health, Dr. Wang shared SNAC’s experience in using federated learning for AI software development and clinical deployment, and discussed the potential of federated learning in the development of artificial intelligence in the medical imaging field.
The collaboration between NVIDIA and SNAC dates back to 2017, when they developed a quantifiable neuroimaging analysis for clinical applications. In 2020, they officially partnered to develop the TRANSCEND federated learning platform, a multi-center artificial intelligence medical imaging platform under the leadership of the University of Sydney. The partnership aimed to combine SNAC’s professional medical image analysis methods and clinical embedding technology with NVIDIA’s NVFlair technology to deploy a federated learning network architecture that meets local Australian needs. Over the next few years, NVIDIA and SNAC worked together to develop and optimize the TRANSCEND platform and promote it to third-party applications. In late 2021, they jointly organized a seminar to showcase the platform’s advantages and achievements in medical image analysis and deployment architecture, which was attended by hundreds of professionals from the medical image analysis field.
Dr. Chenyu Wang emphasized the potential of federated learning in improving the generalization ability and accuracy of models through the increase of data volume and the promotion of direct collaboration among data owners. Federated learning technology is a feasible solution to the challenges in the development of medical image artificial intelligence and is one of the important frameworks to achieve the sustainable development goals of AI medical imaging.