The promise of AI to provide better patient care through accelerated workflows and increased diagnostic capabilities was in full display at RSNA. Catch up with all the news and highlights from the event.
Key Takeaways from RSNA 2018
Empowering Radiology with AI
RADIOLOGICAL SOCIETY OF NORTH AMERICA
RSNA hosted their 104th Scientific Assembly and Annual
Meeting, the largest radiological meeting in the world,
bringing together a diverse community of key luminaries,
industry thought leaders, leading researchers and
premier medical imaging startups.
RSNA: BY THE NUMBERS
80+ Startups, including 41 Inception
431 Session talks
44 Theatre presentations
951 Scientific posters
22 Instructor-Led DLI classes
1,300+ DLI attendees
For over 100 years, RSNA has hosted the world’s
brightest minds in radiology, bringing together key
opinion leaders and industry experts from around
the world to discuss the latest medical imaging
research, new diagnostic tools and protocols, and
NVIDIA participated once again, announcing new
technologies and partnerships, spotlighting
innovative research, engaging in talks and panels,
and offering 22 NVIDIA Deep Learning Institute
(DLI) instructor-led training sessions.
NOW AVAILABLE: NVIDIA CLARA SDK
NVIDIA Clara SDK bundles NVIDIA technology
and expertise specifically for medical imaging
application developers with GPU-accelerated
libraries for computing, visualization, and AI.
The stack can be used at any level to create
world-class applications, from reconstruction,
image processing and rendering, and
computational workflows for CT, MRI and
APPLY FOR ACCESS
ACCELERATING INNOVATION IN MEDICAL IMAGING
NVIDIA announced the Transfer Learning
Toolkit and AI Assisted Annotation SDK for
The NVIDIA Transfer Learning Toolkit enables
developers to harness NVIDIA’s pre-trained
models to fine tune and retrain models using
their own data.
NVIDIA’s AI Assisted Annotation SDK speeds up
the 3D annotation process for CT and MRI
volumes and assists in discovering
abnormalities much faster.
Available early 2019.
RAPIDS FOR HEALTHCARE
At RSNA, NVIDIA showcased RAPIDS, a
new open source project that brings
GPU acceleration to data science
workflows. Applied specifically to
healthcare, RAPIDS demonstrates its
potential to improve clinical care,
operational efficiency, speed up drug
discovery, and advance precision
GET STARTED WITH RAPIDS
COLLABORATING TO BRING AI TO THE CLINIC
NVIDIA also announced partnerships with OSU
Wexner Medical Center and the National
Institutes of Health.
OSU Wexner Medical Center and NVIDIA aim to
build the first in-house AI marketplace on the
NVIDIA Clara platform.
NVIDIA’s partnership with the National
Institutes of Health will focus on bringing AI
tools to clinical trials.
READ THE ARTICLE
HARNESSING THE NVIDIA CLARA AI INFERENCE ENGINE
Watch OSU Wexner Medical Center demo the NVIDIA
Clara AI Inference Engine on their Coronary Artery
Disease and Femur Fracture Models.
SPOTLIGHT ON NVIDIA RESEARCH
3D Brain MRI Brain Tumor Segmentation Using Autoencoder
Automated segmentation of brain tumors provides faster,
accurate diagnosis to enable better treatment plans.
Medical Image Synthesis for Data Augmentation and
Anonymization Using Generative Adversarial Networks
For the first time, researchers are using GANs to generate
abnormal brain MRI images.
Learning Image Restoration without Clean Data, Enhance
Grainy Photos with AI
A deep learning based method that can fix images by simply
looking at examples of corrupted images only.
WATCH THE VIDEOS
AI’S PATH TO THE CLINIC
Kimberly Powell delivered a talk on the technology
breakthroughs that have allowed for the development
of AI tools for radiologists and have created a path to
bring AI to the clinic.
“Through listening to radiologists and our clinical
partners we have learned and created a domain
specific solution for healthcare to accelerate the path
of AI to the clinic.”
-Kimberly Powell, VP of Healthcare, NVIDIA
BUILDING AN AI FUTURE
NVIDIA’s own Abdul Hamid Halabi, Global Business
Development Lead for Healthcare and Life Sciences,
participated in two panels and outlined the three key
pieces needed to make AI a reality in the clinic: Gather
data sets, develop AI infrastructure, and create
algorithms to support and capture the vast amount of
health sensor data.
NVIDIA DEEP LEARNING INSTITUTE (DLI) AT RSNA
RSNA Deep Learning Classroom hosted 1,300+
attendees for 22 hands-on training sessions
throughout the week. Led by NVIDIA DLI
instructors, these sessions focused on how to apply
AI to solve challenging problems in medical
imaging, including 3D Segmentation of Brain MR
and Advanced Data Augmentation Using GANs.
LEARN MORE ABOUT DLI
A GROWING STARTUP ECOSYSTEM
In just one year, startup participation at RSNA
doubled--bringing in over 80 startups, including
41 NVIDIA Inception program partners.
13 Inception partners showcased their cutting-
edge technology in medical imaging in the NVIDIA
booth, including: Zebra Medical Vision, Imagia, 12
Sigma Technologies, IMFusion, Cephasonics
Ultrasound Solutions, EMTensor, Quibim, Image
Biopsy Lab, CureMetrix, Balzano, Subtle Medical,
HeartVista, and NE Scientific.
LEARN MORE ABOUT INCEPTION
NUANCE HOSTS NVIDIA CLARA PLATFORM
Nuance hosted NVIDIA in their booth to show how the
NVIDIA deep learning platform, combined with Nuance’s
PowerScribe radiology reporting and PowerShare image
exchange network, enables development and deployment
of imaging AI models into the AI Marketplace.
“The NVIDIA Clara computing platform powers scalable
inferencing of applications to the Nuance AI Marketplace.
NVIDIA’s AI computing platform is available everywhere,
which gives the AI Marketplace maximum flexibility,
allowing health systems to keep their data securely on
premise, or to take advantage of AI computing in the
-Karen Holtzberger, Vice President and General Manager of
Healthcare Diagnostics, Nuance
To learn more about NVIDIA in