NVIDIA has been turning heads recently in the medical imaging community with their new artificial intelligence-based healthcare programs. Their latest venture in Graphics Processing Units (GPU) technology is Project Clara, a “medical imaging supercomputer” that automates updates on medical imaging devices.
Researchers at NVIDIA were struck by the dichotomy of a rapidly growing medical imaging market and lag time of computation updates. “…It would take decades to update this install base,” said NVIDIA’s vice president of healthcare, Kimberly Powell. Project Clara, in turn, aims to keep imaging instruments and their applications synchronized with pre-existing GPU containers and CUDA libraries.
“A recent algorithm called V-Net uses 3D volumetric segmentation and can automatically measure the volume of blood flowing through the heart,” said Powell. “Fifteen years ago, this algorithm would’ve needed a computer that cost $10 million and consumed 500 kW of power. Today, it can run on a few Tesla V100 GPUs.”
Project Clara is a remote platform that allows various users to access different computational instruments at the same time. It operates on Kubernetes scalability, and can be used with CT, MRI, ultrasound, x-ray, and mammography devices.
To bring Project Clara to life, NVIDIA is partnering with several healthcare companies and hospitals including Martinos Center for Biomedical Imaging’s AutoMap and Subtle Medical who are both developing programs to streamline and improve medical imaging.