• AI
  • Molecular Imaging
  • CT
  • X-Ray
  • Ultrasound
  • MRI
  • Facility Management
  • Mammography

SIIM16: Building Bridges Across Big Threats and Big Opportunities


SIIM16 shed light on patient-centered radiology, artificial intelligence, and their confluence

The Opening Session - Wake-up Call for Patient-Centered Radiology
The SIIM 2016 Opening General Session was presented by Rasu B. Shrestha, MD, MBA, chief innovation officer, University of Pittsburgh Medical Center and President, UPMC Enterprises.

The talk took the top-down approach of putting medical imaging and radiology in the context of significant macro-level changes taking place in populations, technologies, and the health care system. These changes, that UPMC is capitalizing on, highlight growing opportunities for new care models, new technologies and patient-centered care.

More than ever today, this health care transformation stresses the imperative to spur changes in imaging, both incremental as well as paradigm-changing, to move toward patient-centric radiology and value-based imaging.

Having lived through a century of “analog” radiology, followed by several decades of “digital” radiology, the next phase that lies ahead of medical imaging will be the phase of interoperability, analytics, and population health; one where “Context is King.”

In order to navigate into this next phase, medical imaging will need go through a number of transitions: from using technologies designed for regulation, to ones designed for empowerment; from having radiologists constantly at the edge of burnout, to taking joy in their everyday work; from continually adding to the complexity of workflows and processes, to spurring a new wave of simplification; from increasing bureaucracy, to a higher degree of meritocracy; from being merely report-generators, to becoming real physician consultants; from being interpretation-centric to being truly outcomes-centric.

In a nutshell, this next big phase will have medical imaging move from “doing” digital to “being” digital.

The Closing Session - Reality Check on Deep Learning in Medical Imaging
Two days later in the Closing General Session & 2016 Dwyer Lecture, Eliot L. Siegel, MD, FSIIM, professor of radiology, University of Maryland School of Medicine and chief, Imaging Services, VA Maryland Health Care System, gave the SIIM audience another eye-opening talk.

The gist of the session was that medical imaging may well be the ultimate frontier for Artificial Intelligence (AI), yet there is still a ways to go. AI may be winning against world champions at Jeopardy!, Chess or at Go, but it will be long before we can see AI ‘beat’ radiologists, the imaging world champions. If computer-aided detection and diagnosis (CADx) never really took off in a big way, it is for good reason.

In fact, while it is often labeled as AI, deep learning is really only a sub-part of it; and the reality in medical imaging is that we are still at the stage of developing a collection of narrow, discrete deep learning applications –by modality, procedure, and use case, and using them in a supporting or managing function.[[{"type":"media","view_mode":"media_crop","fid":"50063","attributes":{"alt":"Nadim Michel Daher, Industry Principal, Medical Imaging and Imaging Informatics, Frost & Sullivan","class":"media-image media-image-right","id":"media_crop_619351682534","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"6091","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"height: 250px; width: 200px; border-width: 0px; border-style: solid; margin: 1px; float: right;","title":"Nadim Michel Daher, Industry Principal, Medical Imaging and Imaging Informatics, Frost & Sullivan","typeof":"foaf:Image"}}]]

This will likely remain the case at least through the next decade or two, when we may be able to start seeing more general capabilities and more holistic uses of cognitive computing.

Overall, the talk would have comforted those who consider these up-and-coming technologies as a threat rather than an opportunity. It rationalized the fact that it will be many years before these technologies do make the leap forward from the current set of ‘narrow’ and ‘weak’ AI algorithms, to a more general, broad-based, human-level AI that may only then start to threaten to ‘replace’ radiologists.

Perception Change from AI to IA
Both speakers agreed that these technology advances are not about “man versus machine,” rather, about “man versus man + machine.” The misconception is due to the fact that right now, deep learning in medical imaging is right there at, or very close to what Gartner calls the “peak of inflated expectations.”

The next leap can only happen once deep learning in medical imaging has also been taught to leverage omics data (radiomics, genomics, proteomics), and to comprehend a patient’s context at a broad level (natural language processing, structured reports, patient records, wearables data).

The radiology community has yet to come to terms with the looming threat of radiologist replacement, and embrace the technology with the confidence that it can make them better at what they do. This paradigm shift in perception from Artificial Intelligence to Intelligence Amplification (IA) is one that many industries have undergone over the last few years, and that radiology will likely realize over the next few years. As futuristic as it may seem, this is the evolution that will fast-track radiology into the era of cybernetics.

Tying it All Together
Medical imaging has decisively embarked on the journey to value-based imaging. The train has left the station. There too, breakthrough advances, unimaginable only a few years back, are underway.

So, where do these two potentially massive transformations, one in why we practice radiology and one in how we do it, meet at a crossroad?

The answer is elementary, dear Watson: the IA-empowered radiologist will gain more automated tools, deeper insight into their patients, higher confidence in piloting the patient care pathway – essentially more time and means on hand to deliver a truly patient-centered radiology service.

Brought together by Richard H. Wiggins III, MD, FSIIM, CIIP University of Utah Health Sciences Center and newly appointed SIIM Chair, Shrestha and  Siegel deep-dived into two of the most interesting developments going on in medical imaging.

But even more interestingly, they were able to shed light on the big promise that lies at their confluence. On behalf of fellow SIIM attendees and the imaging community at large, thank you for tying it all up together into one cohesive, motivating take-home message.

Related Videos
Expediting the Management of Incidental Pulmonary Emboli on CT
Related Content
© 2024 MJH Life Sciences

All rights reserved.