Nvidia launches cloud APIs to accelerate adoption of AI in medical imaging

AI company Nvidia is launching a new set of cloud-based application programming interfaces (APIs) designed to accelerate the creation and deployment of specialized AI models in medical imaging. The company made the announcement this week at RSNA 2023, the annual radiology and medical imaging conference in Chicago.

Nvidia's new offering is a cloud native extension of the Monai framework. Monai, which stands for Medical Open Network for AI, is the company's open source framework for artificial intelligence in medical imaging.

Monai's primary goal is to make it easier for developers and platform providers to integrate AI into their medical imaging offerings through pre-trained base models and healthcare-specific AI workflows. Over the past few years, vendors have faced some challenges in deploying AI and cloud computing tools - integrating AI in healthcare at scale requires the collaboration of thousands of neural networks, and the industry has shown that it's not quite ready for that.

At the center of the new cloud APIs is Nvidia's VISTA-3D (Vision Imaging Segmentation and Annotation) base model. This model has been trained on a dataset of annotated images from 3D CT scans of more than 4,000 patients with various diseases and body parts. VISTA-3D is designed to accelerate the creation of 3D segmentation masks for medical image analysis, as well as allow developers to refine their AI models based on new data and user feedback.

David Nivolny, director of business development for Nvidia's healthcare business, said in an interview that these new APIs have promising potential to accelerate AI developers' work in imaging. When asked about what kind of AI tools could help Nvidia bring the new APIs to market, he said developers will likely start building models for things like image segmentation before diving into building clinical decision support solutions.

Segmentation involves dividing an image into meaningful regions, which is particularly useful in medical imaging for identifying and delineating structures or abnormalities. Developers can use Nvidia's new APIs to create artificial intelligence models to segment organs, tumors, and other structures in medical images, which can help physicians in diagnosis, treatment planning, and disease monitoring.

In the future, developers will be able to use the API for the most complex tasks, such as classifying diseases. For example, AI developers can use the API to create classification models to identify specific diseases or conditions in medical images - for example, classifying X-rays to detect pneumonia or mammograms to screen for breast cancer.

Creating efficient and cost-effective AI tools for medical imaging requires a domain-specific development framework, Nivolni noted.

"The bottom line is that these new APIs give the healthcare developer community a very powerful set of tools - powered by the Monai community - to build, deploy and scale these AI applications directly in the cloud. That cloud data piece is really a key foundational element. Everything is in the cloud now, even these AI development tools," he said.

Flywheel, a medical image data processing and artificial intelligence platform, has already started using Nvidia's new cloud APIs. Other companies, including medical image annotation company RedBrick AI and machine learning platform Dataiku, plan to migrate to the new offerings soon.

Nvidia wasn't the only company to announce at RSNA 2023 a new offering designed to accelerate the adoption of generative AI solutions in medical imaging. AI startup Hoppr announced a partnership with AWS to launch Grace, a B2B model to help app developers create better AI solutions for medical imaging.