Eight Key Takeaways from RSNA 2021

Reading Time: 5 minutes read

Reflections on AI, COVID and the power of human contact.

Giovanni Lorenz is Deputy, Radiology Product Line at the Defense Health Agency.

As in years past, I left the RSNA conference brimming with deeper thoughts, new ideas and renewed energy. Here are my 8 key takeaways from RSNA 2021. What are yours?  Please comment below.

1. Customers want modalities that incorporate AI
An AI win customers will be looking for in 2022 is some baseline functions. It’s apparent that we’re all excited about the AI space and know that the current valuation for specific smart and machine learned workflows is just starting to get notice. There are large unresolved problems where AI could help. These include silos, subspecialty knowledge, workflow efficiency, prioritization and smart worklists, and enterprise presence to leverage slack in certain areas of the reading population. Modalities that incorporate AI to observe, direct and flag exams should be a baseline 2022 offering.

2. COVID-19 was so last year

The real problems are bigger than variants and will require extra radiologic leadership and clinician support. This includes operational elasticity to meet surge; enterprise under cloud and blockchain considerations; guideline and screening program development, and more. At the moment, none of this appears to be tied directly to reimbursement.

Also, we learned during the pandemic that demand can be met- like a combustion engine running through fuel. But we could be more efficient by leveraging AI for extraction and report assistance. 

3. DR is not dead!

Most of my (and probably your) reading list is radiography. And that’s a good thing! We want services to be easily ordered and accessible (and the list proves it). Let’s be ready to embrace new innovations like flexible detectors, cloud-based distribution of workflow, and modality AI. 

4. We have a payment and leadership problem

Radiologists and departments are leading the acceptance and usage of technological advancement – as usual. However, purchasing power and clinical acceptance has to consider our downstream providers and patients (and population health experts). Where are the inter-conference and department discussions? How many institutions have set these up to be more strategic about yielding smarter workflow and learned machines?

5. How should we treat AI?

I personally think we should treat AI as a human resource. Simply put, this is a human resource and should be looked at with consideration as an FTE.  In a panelist discussion at RSNA, I suggested we treat it as O/M budget (like CT) or something completely different. Perhaps as a hospital expense based out of ethical or safety controls? The discussions need to commence… it’s not going away.

6. In search of open platforms

I think modalities coupled with AI solutions could bridge some of the available fringe software applications; and I hope to eventually see open platform and unified software protocol agreements. I’d certainly like to see uniform protocols of AI usage (detection, efficiency, clinical decision support, and security) that could lead to eventual payments by CMS. 

7. Inter-vendor collaboration

“Collaboration” and “democratization” were ways to describe the industry’s attempt to adjust to a shrinking labor and solutions pool. A likely roadmap to support the short-term gains needed by the industry could be offering a larger variety of inter-vendor collaboration to see which services (SaaS, etc) customers are looking to add for 2022.

8. Human interaction is good for the soul

Nothing can convince me or the 22,000-plus who showed up to RSNA21 that the physical presence of a meeting was wasted. Relationships develop cause, which lead to success. So I enjoyed that slice at Connie’s, shared my table at lunch, met new and wise friends whom I knew from countless Microsoft Teams Chats…and enjoyed shaking their hands.

Interacting with a face on a screen is far short of human contact. From the Rad technologist at the bedside to the Radiologist reading the exam-we all need it. While I’m excited for the potential of non-essential metaverse applications (and there are several we can discuss later), we still need a doctor-patient visit. Contact with each other and with our patients should continue to be a goal for Radiologists and all of us within the imaging industry.  

Lastly, I want to give a shout out to all the RSNA staff, educators and vendors who made the conference a success.

I hope you enjoyed reading my 8 key takeaways from RSNA 2021. I’d love to hear your insights, contradictions or anything else you have to add. Post them in the comments section below. And I’ll see you next year at RSNA!

Giovanni Lorenz is Deputy, Radiology Product Line for the Defense Health Agency at San Antonio Military Health System. He is a double board certified Cardiothoracic Radiologist with 20 years of active duty experience within the Air Force. Opinions and recommendations are personal and do not reflect the USAF, DoD/DHA or US government. 

#RSNA21 #RSNA2021