• 8:30 AM - Registration
• 9:15 AM – Introduction
• Clara Platform 1.0 GA Overview
• Medical Imaging AI brief
• NGC & NGC-ready systems
• 10:00 AM – Coffee Break
• 10:30 AM - Clara Train – AIAA

• AI Assisted Annotation (AIAA) Tool

• Live Demo of Clara Train SDK
• 11:00 AM - The Medical Imaging Interaction Toolkit (MIKT) Demo
• MITK toolkit related workflow and capabilities live demo
• Annotation workflow live demo
• 11:15 AM - Clara Train – TLT
• Transfer learning toolkit for Clara Train
• Live demo of Transfer learning workflow
• 12:00 PM – Lunch
• 1:00 PM – Clara Deploy
• Live demonstration of Clara Deploy SDK
• 1:45 PM – TensorRT
• Introduction to TensorRT and TRTIS
• INT8 calibration and its peculiarities in the medical domain
• 2:30 PM - Clara Train Demo by SNAC
• Demonstration by SNAC team on Clara Train and Deploy for Neuroimaging use case
• Q&A
• 3:00 PM – Coffee Break
• Open discussion between participants
• Q&A
• 4:00 PM – End of workshop

Instructor will run live Clara demo concentrating on explaining all the working parts. If you would like to run excercises on your machine, please bring a fully charged laptop and do the following preparation prior to the workshop. Running hands-on is optional and not compulsory to benefit from the workshop.

Hands-on preparation:

1. Download the docker version for:

• Clara Train at NGC (nvcr.io/nvidia/clara-train-sdk:v1.0-py3),
• Clara Deploy at NGC (https://ngc.nvidia.com/catalog/model-scripts/nvidia:clara_deploy_sdk)

1. Install the latest Nvidia Driver for your GPU in linux environment (e.g 410+) and Compute Capability 5.2 and above
2. Install Cuda version 10.1 on your laptop
3. Download the data: https://drive.google.com/open?id=1A2IU8Sgea1h3fYLpYtFb2v7NYdMjvEhU

We will share access to (1) Jupyter notebook for hands-on exercise and (2) presentation materials after the workshop.

Workshop Instructor: Dr Alvin Ihsani, Deep Learning Solutions Architect, Medical Imaging, NVIDIA Corporation

In collaboration with: