Today, GPUs are found in almost all imaging modalities, including CT, MRI, x-ray, and ultrasound - bringing compute capabilities to the edge devices. With the boom of deep learning research in medical imaging, more efficient and improved approaches are being developed to enable AI-assisted workflows.

To develop these AI capable applications, the data needs to be made AI-ready. NVIDIA Clara’s AI-Assisted Annotation does so by providing APIs and a toolkit to bring AI-assisted annotation capabilities to any medical viewer. Post annotation, data scientists and researchers need to build a robust AI model. To enable this, NVIDIA Clara Train includes techniques like AutoML, privacy-preserving federated learning and Transfer Learning. One trained AI model is available in Clara Deploy and provides a reference framework to take an AI model and write an application workflow around it to enable interfacing in a hospital like environment. Clara Deploy includes platform capabilities required to support multi-AI, multi-domain workflows in a seamless manner.

Download Clara Train      Download Clara Deploy

New Clara Features to Help COVID-19 Research

Providing researchers an additional tool in the battle against COVID-19, NVIDIA Clara Imaging has released Clara Train MMARs for CT Lung segmentation and COVID-19 Chest CT Classification, providing researchers with a state-of-the art implementation and ability to optimize these models with the Clara Train application framework. Also available is the Clara Deploy reference pipeline for COVID-19 Chest CT Classification to provide researchers with a reference deployment pipeline which can be seamlessly evaluated for localized data.

Download COVID-19 Models

Running on a variety of hardware solutions, the Clara application framework leverages the edge stack
to deliver all the necessary tools for healthcare application developers.

The Clara Train reference applications is a domain optimized developer application framework that includes APIs for AI-Assisted Annotation, making any medical viewer AI capable and a TensorFlow based training framework with pre-trained models to kick start AI development with techniques like transfer learning, federated learning and AutoML.

Clara Train includes AI-Assisted Annotation APIs and Annotation server that can be seamlessly integrated into any medical viewer making them AI capable. The training framework includes decentralized learning techniques like federated learning and transfer learning. Clara Train also makes available model applications packaged as MMARS (Medical Model ARchive) available to users, providing an intuitive config based environment for data scientists and researchers to get kick-started with AI development.

NVIDIA’s latest release of Clara Train includes the AutoML module. Clara Train’s AutoML functionality makes the process of hyper-parameter tuning seamless by intelligently searching for optimal parameter settings to train models automatically.

Read our developer blog to learn more about Clara Train’s AutoML feature.

Learn More

New AI-Assisted Annotation Features

  • Interactive annotation kick-started with new AI Model for unseen organs allowing for image labeling when a pre-trained model is not available
  • Customize annotation workflows by bringing your own models or models to an annotation workflow.

Horovod - Automatic Mixed Precision - Smart Cache - 8 GPUs
Click on graph for more results

New Training Framework Features

  • AI models for chest CT for COVID-19 detection.
  • Clara Train Federated learning v3.0 enables researchers to collaborate and build AI models without sharing private data. This feature is further enhanced by allowing users to extend this reference implementation by bringing their own component implementations like aggregators and privacy-preserving policies.
  • AutoML makes the tedious and time-consuming task of parameter-tuning and testing for a new model seamless by searching the most optimal parameters to apply in a training workflow. The system enables researchers to bring their own AutoML component implementation for Controller, Hander and Executor.
  • Performance optimizations - Up to 50X faster training with domain-optimized performance features like TensorFlow Keras based imaging pipeline, Smart Cache, Nova Grad optimizer, Automatic Mixed Precision etc.

Running NVIDIA Clara Train on AWS Cloud

Clara Train runs on-premise and on the cloud is now deployable in a highly available (HA) configuration on the AWS Cloud. The AWS Quick Start lowers the barrier to bringing up a secure, scalable infrastructure on the cloud. This deployment provides scalable access to NVIDIA V100 Tensor Core graphics processing units (GPUs) and the Amazon Elastic Compute Cloud (Amazon EC2) P3 instance type. This deployment is based on Amazon Elastic Container Service (Amazon ECS) and Amazon EC2. Amazon Elastic File System (Amazon EFS) is used for storage shared between containers.

Get Started on AWS Read Blog

An extensible reference development framework that facilitates turning AI models into AI-powered clinical workflows with built-in support for DICOM communication and the ability to interface with existing hospital infrastructures.

Clara Deploy provides a container-based development & deployment framework for mutli-AI, multi-domain workflows in smart hospitals - One platform managing and scaling Imaging, Genomics and Video Processing workloads. It uses Kubernetes under the hood and enables developers and data scientists to define a multi-staged container-based pipeline.

NVIDIA’s latest release of Clara Deploy SDK includes support for multi-AI, multi-domain workflows - one architecture orchestrating and scaling imaging, genomics, and video processing workloads. This reference platform delivers a unified foundation to enable intelligent workloads in smart hospitals.

Read our developer blog to learn all of Clara Deploy’s latest features and reference applications.

Learn More

New Platform Features

  • Supported pipeline composition with strongly typed operator interface enabling pre-runtime validation of pipelines, compatibility of concatenated operators in terms of data type, and allocation of memory for the pipeline using FAST I/O via CPDriver.
  • Built-in scheduler to help manage and allocate resources to execute pipeline job.
  • An AI model management tool with the ability to store and manage models locally through user inputs, pull models in from external stores via direct download services to create and manage model catalogs.
  • A CLI Load Generator to simulate expected hospital workloads to ensure both the architecting of hardware and software can support the development of application pipelines.
  • A Fast I/O feature to provide an interface to memory resources that are accessible by all operators running in the same pipeline. Memory resources can be used for efficient, zero-copy sharing and passing of data between operators.

New Reference Application Pipelines

  • Detection of COVID-19 in CT datasets
  • Digital Pathology
  • Usage of Shared Memory in Multi AI CT Pipeline
  • DICOM Series Selection Pipeline
  • COVID-19 Classification for X-ray
  • Prostate Segmentation
  • Multi-AI Organ Segmentation
  • 3D Cropping using shared memory
  • DeepStream Video Batch

By clicking on the download Clara Deploy, you agree to the terms and conditions in this end user license agreement.

This is a release candidate for R5_3 - we will be providing hotfixes soon - please share any issues on dev forums.

To enable a whole new world of devices, Clara AGX enables medical devices that need the ability to perform real-time AI and advanced image, video, and signal processing.

NVIDIA Clara AGX™ Developer Kit

The NVIDIA Clara AGX™ developer kit is an embedded AI computer and software development framework for medical devices that need the ability to perform real-time AI and advanced image, video and signal processing. NVIDIA Clara AGX securely manages and orchestrates AI application deployments to fleets of medical devices or edge nodes.

Coming Soon

Other Developer Tools

Clara reference applications are built on CUDA-X, a collection of libraries, tools, and technologies that deliver dramatically higher performance than alternatives across multiple application domains—from artificial intelligence to high-performance computing. Commonly used tools and APIs are provided here for your convenience.


CUDNN TensorRT Triton Inference Server


Image and Signal Processing


OptiX Video Codec Deepstream SDK Optical Flow


Webinars      Blogs

deep learning Get Started With Hands-On Training

The NVIDIA Deep Learning Institute (DLI) offers hands-on training for developers, data scientists, and researchers in AI and accelerated computing. Start your hands-on training in AI for Game Development with self-paced courses in Computer Vision, CUDA/C++, and CUDA Python. Plus, check out two-hour electives on Deep Learning for Digital Content Creation and Game Development.

Disclaimer: Clara and samples are for developmental purposes only and cannot be used directly for clinical procedures.