Skip to content

Machine Learning in Radiology - Takeaways from RSNA 2018

Machine Learning in Radiology - Takeaways from RSNA 2018

At RSNA 2018 in Chicago last week, 150 companies were demonstrating some form of AI technology according to one estimate. I was left with the impression that early commercial adoption of the latest wave of AI/ML diagnostic technologies has begun and start-ups in this space need to begin factoring Asia into their global expansion plans.

At the RSNA Machine Learning Showcase, several Asia-based companies were prominent. China had a strong presence with 10+ AI algorithm start-ups exhibiting (Yitu Health, InferVision, 12 Sigma, Deepwise, Curacloud, Huiyihuiying, LinkingMed, Synyi AI, etc out of ~40 China-based imaging AI start-ups according to this report), followed by Korea with three start-ups (Lunit, DeepNoid, VUNO), while Japan had one (LPixel) and was represented otherwise by large imaging equipment and software vendors.  

I spoke with several interesting AI algorithm 'platforms' aiming to provide routes to market for algorithm developers. Terarecon’s EnvoyAI seemed to have the most traction (28 companies on their Exchange, offering 69 AI apps of which 17 were FDA cleared), followed by Blackford Analysis (with 9/9 FDA cleared algorithms), Qmenta (3/0 FDA cleared), and others. Algorithm developers are also partnering with PACS and imaging equipment vendors to gain access to customers and data. 

On the regulatory side, the US FDA has approved the largest number of AI-aided diagnostic products via the July 2017 Digital Health Innovation Action Plan making the US the initial market of choice for many algorithm developers, including those based outside the US. China’s NMPA(CFDA) has approved at least two AI diagnostic products and has just updated its medical device catalogue as of August 2018. Korea's MFDS has approved at least one product, while the Japan PMDA is in the process of developing regulations to accommodate new AI applications.

The takeaways for me were: opportunities exist for innovative AI algorithm developers to partner with platforms, software and hardware companies to access global markets - Japan in particular appears interesting with fewer 'native' AI algorithm start-ups but multiple mature imaging hardware and software vendors (FujiFilm, Canon, Konica Minolta, etc). Evolving China and Korea-based AI platforms (Infervision, etc) and equipment vendors also present interesting partnership opportunities in Asia.

Sol Bergen-Bartel
About the Author

Sol Bergen-Bartel

Sol has worked in and with businesses in Asia for 18 years and leads Intralink’s Medical & Life Science business development work in North America. He has a degree in molecular biology from the University of Virginia and spent 10 years based in Intralink’s Shanghai office including as VP of Operations. Sol has driven Asia expansion for companies across life sciences, diagnostics, medtech, semiconductors, renewables and other industries, closing deals and forming partnerships in China, Japan and South Korea. He is fluent in Mandarin and based in our Boston office.

We use cookies to give you the best experience of using this website. By continuing to use this site, you accept our use of cookies. Please read our Cookie Policy for more information.