Peter Zhang
Oct 08, 2024 19:36
NVIDIA’s AI tools, including NIM microservices, are transforming the U.S. healthcare landscape, enhancing research and clinical practices with cutting-edge AI technology.
The U.S. healthcare sector is experiencing a technological transformation as digital health agents, powered by artificial intelligence, are being integrated across various facets, from research labs to clinical environments. This trend was highlighted during the NVIDIA AI Summit in Washington, D.C., where NVIDIA unveiled its suite of AI-accelerated tools designed to enhance healthcare capabilities.
NVIDIA’s AI Innovations
Among the innovations showcased were NVIDIA NIM, a collection of cloud-native microservices that facilitate AI model deployment and execution, and NIM Agent Blueprints, a catalog of customizable, pretrained workflows. These technologies are currently being utilized in the public sector to advance medical imaging analysis, aid therapeutic research, and extract data from extensive PDF databases.
In a practical application, researchers at the National Cancer Institute, part of the National Institutes of Health (NIH), are leveraging AI models developed with NVIDIA MONAI for medical imaging. These models include the VISTA-3D NIM foundation model, which assists in segmenting and annotating 3D CT images. Additionally, the NIH’s National Center for Advancing Translational Sciences (NCATS) employs the NIM Agent Blueprint for generative AI-based virtual screening, aiming to streamline the drug development process.
Wider Applications and Availability
NVIDIA’s NIM microservices and blueprints are accessible via various cloud service providers and technology solutions, offering medical researchers across the public sector the opportunity to adopt state-of-the-art AI models. The tools are designed to be customizable with organizational data and can be refined with user feedback.
Developers can access numerous NIM microservices and blueprints, which can be deployed in production using the NVIDIA AI Enterprise platform. Key blueprints include those for generative virtual screening in drug discovery, multimodal PDF data extraction, and the creation of interactive AI-powered avatars for customer service, applicable in telehealth and patient care.
Enhancing Drug Discovery
NCATS is actively evaluating the NIM Agent Blueprint for virtual screening, utilizing NVIDIA’s RAPIDS software libraries to accelerate drug discovery research. By employing GPU-accelerated data science tools like the cuGraph and cuDF libraries, researchers can efficiently map chemical reactions, significantly reducing processing times.
The Genetic and Rare Diseases Information Center, managed by NCATS, is exploring AI tools for PDF data extraction to enhance information retrieval from complex databases, aiding in the response to queries from individuals with rare diseases.
Collaborations and Future Prospects
Startups and global integrators are incorporating NVIDIA’s AI solutions into their platforms. A notable example is Abridge, an NVIDIA Inception startup awarded a contract by the U.S. Department of Veterans Affairs to streamline clinical documentation through AI-powered transcription services. Additionally, AWS HealthOmics and ConcertAI are leveraging NVIDIA’s AI capabilities to advance drug discovery and oncology AI technology, respectively.
Deloitte, a global systems integrator, is deploying the NIM Agent Blueprint for virtual screening to assist federal health agencies in adopting AI solutions for drug discovery. This collaboration underscores NVIDIA’s commitment to making AI tools widely accessible within the biomedical research community.
For more on NVIDIA’s healthcare AI innovations, visit the NVIDIA Blog.
Image source: Shutterstock
