April 27, 2018 — Computer technology company NVIDIA and Canon Medical Systems announced a new partnership to develop the research infrastructure for enhanced artificial intelligence (AI) in healthcare. The partners hope to make a significant contribution to promoting the use of data-intensive deep learning techniques in medical and related research, as well as to drive the uptake of AI in the healthcare sector.
According to NVIDIA, the healthcare sector needs to analyze scientific reports from around the world, while simultaneously coordinating a variety of patient data to determine the most appropriate treatment options. Given the huge volumes of data involved, big data analysis via deep learning will play a major role in the development of optimized healthcare delivery systems and support early detection and assisted diagnosis.
At the same time, medical institutions wanting to use deep learning for independent research need hardware for analysis, systems for the collection, collation and analysis of in-house data, and knowledge of deep learning processes and techniques.
Canon Medical Systems will use NVIDIA DGX systems to process large volumes of medical data generated by Abierto VNA (vendor neutral archive), the proprietary, in-house medical data management system it launched in January.
DGX systems feature NVIDIA Tesla data center graphics processing units (GPUs) powered by the Volta advanced GPU architecture. Among the portfolio is NVIDIA DGX Station, an AI workstation that has the computing capacity of four server racks in a desk-friendly package, while consuming only 1/20 the power.
The systems include NVIDIA’s specially optimized AI software and Canon Medical System’s graphical user interface, which provides full support for the design, deployment and operation of advanced deep learning algorithms.
Deep learning usually requires extensive programming and data science skills. However, the system from Canon Medical Systems and NVIDIA guides users through all the steps involved in the deep learning process, from generating training data with the image viewer to setting up the NVIDIA learning environment.
For more information: www.us.medical.canon