NVIDIA Expands Clara Platform with AI and Genomics Innovations to Combat COVID-19 at GTC 2020

On May 15, NVIDIA announced significant expansions to the NVIDIA Clara™ medical platform, including performance enhancements, key partnerships, and new features, aimed at helping the medical community better track, test, and treat COVID-19.

This platform integrates AI and accelerated computing to help medical researchers, technology solution providers, and hospitals respond to the pandemic more quickly and in new ways:

• Breaking the Genome Sequencing Speed Record -Provide COVID-19 researchers with 90 days of free access to NVIDIA Clara Parabricks®. This software set a new speed record by completing whole-human genome DNA sequence analysis in just 20 minutes.

• Disease Detection AI Model- An AI model co-developed with the U.S. National Institutes of Health (NIH) goes live today. This model assists researchers in assessing the severity of COVID-19 through chest CT scans and facilitates the development of new tools for improved viral research, evaluation, and infection detection.

• NVIDIA Clara Guardian, Developed Specifically for Smart Hospitals- NVIDIA Clara Guardian™ was released today. Leveraging intelligent video analytics and automatic speech recognition technologies, this product enables next-generation smart hospitals to monitor patients’ vital signs while minimizing physical contact.

Kimberly Powell, Vice President of Healthcare at NVIDIA, stated, “The COVID-19 pandemic has fostered deeper collaboration among the technology, research, and healthcare sectors to jointly develop new computing solutions, thereby accelerating our understanding of the disease’s transmission patterns, scale, and severity. Never before has there been such an urgent global need to apply cutting-edge AI and accelerated computing technologies across all areas of healthcare, and their impact will be widely felt both in the post-pandemic era and throughout the future of the healthcare industry.”

Running NVIDIA Clara Parabricks on the newly released NVIDIA A100 GPU reduces whole-genome DNA sequencing analysis time to under 20 minutes, breaking previous records. NVIDIA also introduced a GPU-accelerated RNA sequencing solution that delivers results within two hours, providing researchers with critical insights into patients’ disease susceptibility, disease progression, and treatment response.

“Time is critical to our work,” said Glen Otero, Vice President of Scientific Computing at the Translational Genomics Research Institute. “Researchers at the Center for Rare Childhood Disorders are using NVIDIA Clara Parabricks to identify the most effective treatment for each patient and deliver results on the same day.”

Accelerated genomic analysis paves the way for creating larger, higher-quality genomic datasets. The Abu Dhabi Department of Health recently announced a population genomics initiative that is leveraging the Group 42 Artemis supercomputer, powered by NVIDIA DGX™, and NVIDIA Clara Parabricks to build a genomic reference for UAE citizens, with plans to integrate genomic analysis into comprehensive initiatives aimed at improving clinical care.

To enable users to more easily leverage NVIDIA Clara Parabricks for acceleration in the cloud, NVIDIA is partnering with DNAnexus, an enterprise genomics cloud platform, to ensure seamless service delivery, program execution, and result return for users.

NVIDIA also provides researchers with another tool to combat COVID-19—a set of AI models that can help researchers detect and study infected patients using chest CT scan data.

These models are now available in the latest version of Clara Imaging on the NGC™ Software Center.

This suite of models was jointly developed by NVIDIA’s Applied Research team and clinicians and data scientists from the U.S. National Institutes of Health (NIH), using data from regions with high COVID-19 infection rates, including China, Italy, Japan, and the United States.

The NVIDIA Clara application framework for medical imaging was leveraged to build this suite of AI models. NVIDIA Clara includes tools tailored for domain-specific AI training and deployment, enabling NVIDIA and the U.S. National Institutes of Health (NIH) to develop these models in under three weeks.

Building robust AI models is currently the top global priority, yet cross-border data sharing remains a significant challenge. Currently, NVIDIA and Mass General Brigham are leading a new multinational collaborative research initiative on COVID-19, which extends AI models for the novel coronavirus to X-ray imaging. This enables the local deployment of these models without sharing any patient data, thereby safeguarding patient privacy.

The all-new NVIDIA Clara Guardian application framework provides hospitals with a much-needed ecosystem of AI solutions for public safety and patient monitoring by transforming ordinary sensors into smart sensors. Its primary use cases include automated temperature screening, face mask detection, safe social distancing, and remote patient monitoring.

Partners within this ecosystem are leveraging pre-trained models and transfer learning techniques to develop and deploy AI applications that integrate computer vision, speech recognition, and natural language processing. Powered by the NVIDIA EGX™ edge AI platform, solution providers can securely deploy and manage a large number of devices across hospital environments.

Currently, dozens of solution partners—including AnyVision, BriefCam, Care.ai, Chooch AI, Deep Vision, Diycam, IntelliSite, Malong Technologies, Ouva, SafelyYou, SAFR, SmartCow, TeiaCare, Tonbo Imaging, and Whiteboard Coordinator—have deployed NVIDIA Clara Guardian-based solutions in more than 50 hospitals and over 10,000 patient rooms worldwide.