Using innovation and AI to redefine radiology: 5 takeaways from RSNA 2021

By Richard Hausmann, CEO

Back in my home office in Germany, I’m settling in after a week in Chicago for the 2021 Radiological Society of North America (RSNA)’s scientific assembly and annual meeting. In its 107th year, the theme of RSNA 2021 was “Redefining Radiology.” Boasting over 25,000 registrants and more than 2,000 presenters, the conference explored the new technologies and ideas that are reshaping the future of the field. At our booth in the AI Showcase, we joined other exhibitors in displaying current and emerging artificial intelligence (AI) technologies utilizing medical imaging. It was exciting to see and learn about cutting-edge solutions and new perspectives, as well as interact with clinicians, executives, and other vendors.

Here are my 5 takeaways from the event:

1. AI is at the forefront of radiology.

I was impressed by the breadth of AI in the imaging space — the number of vendors and applications continues to grow. Like many industries, the landscape of radiology and the future of clinical practice are being developed and enhanced by AI. This was reflected throughout the program, with sessions exploring:

  • the potential of AI in breast, abdominal, oncology, and emergency imaging;
  • ethics in AI;
  • data sharing and patient privacy; and
  • whether AI is a long-term investment or an immediate opportunity (our opinion: it’s both).

The connecting thread is that AI has the potential to improve the quality and accuracy of medical imaging. For example, while viewing a demo at another company’s booth, we were asked to identify a small bone fracture on an x-ray. Because it is so difficult to see with the human eye alone, we required assistance from their AI technology, which was more sensitive at detecting slight differences in the image. It’s exciting to see what AI makes available to clinicians that they wouldn’t have been able to access previously — and the possibilities that it brings to improving patient care.

2. Larger, more established companies AND start-ups: in the AI space, it’s not either/or.

It’s easy to think of innovation in any space as the giants vs. the little guys: the Davids vs. the Goliaths. When it comes to innovation in medical imaging, however, there is a clear need for both sets of players. The larger vendors are looking at how they can implement AI algorithms in their existing applications, workflows, and hardware to enhance their capabilities, for example, to improve diagnosis when using less precise imaging modalities. And, from these companies, we’re seeing an increase in platforms integrating multiple AI applications as well as proprietary user interfaces that make available numerous third-party algorithms and applications in one location.

While they also develop one or the other own application, the “Goliath” platforms need access to lots of innovative applications, which is where companies like ours have an important role to play. Start-ups or scale-ups are really well-suited to creating specialized applications for clinical use that incorporate AI from the start. Our solutions are based on an AI foundation, using our long-standing scientific expertise and continued search for answers to clinical challenges. Because of our agile, creative team, we have the ability to pivot and develop innovative applications in response to emerging research and drug development, such as new disease-modifying drugs for Alzheimer’s disease.

3. Data is everything.

This is something we hear from the clinicians who we work with regularly and something we address through our applications. We pride ourselves on providing solutions that go beyond imaging and MRI biomarker quantification. This was also a key theme for visitors to our booth. “Are you the guys who have that application…?” they asked, wanting to see a demo of cDSI™, our AI-based decision support tool for patients with neurological disorders.

To provide value for clinical decision making, AI applications need access to validated, high-quality datasets with a broad range of patient information. cDSI has just that — access to a variety of patient data, including imaging biomarkers, clinical information, neuropsychological tests, and more. It compares this information for the current patient to a database of similar information from previous patients, to provide clinicians with deeper insights into their patient’s condition. With these insights, clinicians can make earlier, more confident diagnosis; monitor treatment effects; and even predict disease progression.

4. AI tools are needed to fill the gaps in research and clinical workflows.

MRI quantification is an important component of improving patient care through more consistent, more objective reads, and there are many tools out there that do this well. However, there was great interest in tools that support radiologists, clinicians, and researchers throughout the entire workflow for patient care. We heard from quite a few booth visitors that they were excited to see a product, like cDSI, that helps with early disease detection, differential diagnosis, the most likely disease progression, and long-term monitoring.

We also demonstrated two of our new applications currently in development, cPET™ and cARIA™, that are aiming to address other needs for clinicians to provide better care for their patients as well as researchers to develop new effective and safe treatments. cPET provides fully automated quantification of FDG and amyloid PET tracers, introducing another imaging modality that complements the existing information in cDSI for improved diagnosis and management of dementia and other neurodegenerative diseases.

We developed cARIA in rapid response to the growing need to monitor patients on new and investigational disease-modifying drugs (DMDs) that remove amyloid in conditions such as Alzheimer’s disease. It focuses on longitudinal monitoring of amyloid-related imaging abnormality (ARIA; which presents as new microbleeds or edema) that could occur as a drug side effect. As far as I’m aware, this is the first time that there has been a true need for regular MR follow-up to detect abnormalities in response to drug treatment, and we’re excited that we’ll be able to have this application available for research and clinical use soon.

5. Conferences are still important and relevant to the medical, scientific, and technology communities.

After nearly two years of virtual meetings from the comfort of our homes, I was a bit unsure if there would be a drop in attendance and how that would impact the vigor of the meeting. To my surprise, however, the conference felt well-attended, we had fantastic traffic at our booth, and engagement with our products was high. The conference organizers took great care to ensure the health and safety of everyone in attendance, so we could all feel comfortable interacting with others and let the true purpose of RSNA take center stage: innovation in medical imaging. As a result, we were able to have productive, interesting conversations with people — conversations that underscored the first four themes above.

That leads me to one of my favourite aspects of RSNA, which was the ability to connect with my peers, colleagues, and friends from other companies. It was fantastic to be together in the same space, when normally we are spread throughout the world — Finland, Germany, Great Britain, the US, and more. There are many benefits to working remotely, but when it comes down to it, there is nothing that compares to face-to-face interaction. One thing is for certain — I’m looking forward to the next opportunity to gather with peers, colleagues, and clinicians. In the meantime, there is much to reflect on from RSNA that will inform our innovation going forward.