SAN JOSE, CA - Graphics processing unit (GPU) technology developer NVIDIA has debuted a new initiative, called Project Clara, designed to provide the power of supercomputing to legacy medical imaging scanners installed in the field.
Speaking on March 27 in the keynote address at the company's annual GPU Technology Conference (GTC 2018), NVIDIA President and CEO Jensen Huang discussed Project Clara, which he described as a virtualized data center that would be available on a remote basis for multiple modalities and users.
Project Clara is designed to address the obsolescence of older imaging scanners that can't be updated with newer generations of GPU or supercomputer technology. These older systems have limited access to artificial intelligence (AI) and image processing technologies that can be built into new systems.
Comparing a modern ultrasound scan with one acquired 15 years ago, Huang pointed out the difference between the older scan with its grayscale pixels and the modern, rendered fetus with accurate flesh tones. He then demonstrated how the 15-year-old scan could be streamed to a data center to produce a volumetric, modern view. Applying additional AI technology to the scan can infer details about structures such as the left ventricle of the heart, which was segmented out in 3D.
Huang said NVIDIA is working with dozens of healthcare companies, research institutions, and start-ups on the technology.
