The University of California, San Francisco is employing Nvidia technology to help develop artificial intelligence tools for clinical radiology.

WHY IT MATTERS
The two organizations will work together on several AI projects, including brain tumor segmentation, liver segmentation and clinical deployment, leveraging Nvidia's Clara healthcare toolkit and the tech giant's DGX-2 AI system.

Clara Medical Imaging provides developers with the tools to build, manage and deploy intelligent imaging workflows and instruments, while Clara Genomics addresses the growing size and complexity of genomics sequencing and analysis with accelerated and intelligent computing.

Powered by DGX software and the scalable architecture of Nvidia NVSwitch, the DGX-2 is a 2 petaFLOPS system combining 16 interconnected graphical processing units – the system could help UCSF researchers significantly cut the time to train AI models.

The number of images acquired during common studies such as MRI and CT scans has swelled in recent years corresponding with the growing number of patients being imaged.

The two organizations will also partner on predictive analytics research to help develop tools for radiologists and other physicians for imaging scans and medical records, as well as AI models that can be deployed into the medical center's imaging workflow.

UCSF and Nvidia will also work on AI technology that can be used to de-noise medical images, deep learning algorithms to optimize how the medical center's fleet of imaging scanners is used, and to analyze MRI scans of the brain and liver.

THE LARGER TREND
The announcement comes as a growing number of universities and tech giants are deploying AI to help with medical scanning and imaging technologies.

Last month the FDA cleared GE Healthcare's AI platform for X-ray scans, Critical Care Suite, which was developed in partnership with UCSF and is powered by GE's Edison AI technology.

The platform can help radiologists prioritize cases involving collapsed lungs, while the Edison platform's embedded data processing analytics provide for automated AI quality check features, which can detect acquisition errors.

ON THE RECORD
The number and size of imaging files, which can take up gigabytes or now even terabytes of data storage, have made for an "absolutely overwhelming volume" of information to digest, UCSF's head of the department of radiology and biomedical imaging, Christopher Hess, said in a statement.

"We're hoping to use AI to help radiologists better navigate and interact with data, to derive more meaning out of images, and to improve the value of medical imaging for the individual patient," Hess explained. "We're interested in integrating data from not only imaging, but also from medical records, genetics and other information sources in the healthcare system."
 

Nathan Eddy is a healthcare and technology freelancer based in Berlin.
Email the writer: nathaneddy@gmail.com
Twitter: @dropdeaded209