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

Radiomics Come of Age at RSNA 2015


At RSNA 2015, it was clear that machine learning and radiomics are in radiology’s future.

Most imaging informatics aficionados are likely to agree on one thing: it takes a lot more scouting on the exhibit floors of industry tradeshows than it used to to get the ‘WOW!’ effect.

The show floors of imaging and informatics conferences such as RSNA, HIMSS, or SIIM are bustling with innovation, like they have always been, and RSNA 2015 was no exception. However, whether it is breakthrough or incremental, the innovation of the last few years has definitely carried a higher sense of realism than one of science-fiction-until radiomics came along.

RSNA 2015: Mostly Incremental Advances in Enterprise Imaging Informatics
Continuing, for the most part, on the same track of the last five years, imaging informatics vendors continue to work on stepping up their game in enterprise imaging. Some vendors’ technologies and products are worthy of closer attention and the ‘hottest’ are likely to fall under one of these four product categories:

• Third-generation vendor-neutral archiving (VNA) solutions build on first-generation (departmental, DICOM images) and second-generation (multi-site, multi-departmental DICOM). VNA solutions bring together both DICOM and non-DICOM images, as well as relevant non-image data, into an integrated, standards-based health care content management (HCM) platform facilitating seamless cross-enterprise data sharing and multi-disciplinary collaboration.

RSNA 2015 Vendor Highlights: Watching this new generation of integrated HCM take shape, we, at Frost & Sullivan, continue to be impressed by Lexmark Healthcare’s momentum, are gaining confidence in Hyland OnBase’s unfolding strategy, hold a wait-and-see attitude regarding FujiFilm’s Teramedica integration and Vital Images’ late entry into the game, take notice of BridgeHead’s increased traction stateside, and commend Carestream Health’s major new developments in HCM.

• Cross-enterprise workflow orchestration solutions offer consolidated workflow solutions for the entire imaging service line across multi-site, multi-PACS, multi-vendor environments. Routing the right study to the correct resource based on availability, sub-specialty, and credentials requires much more than merely combining individual worklists into a single ‘mega’ worklist; it takes intelligent and customizable built-in tools to normalize the data, streamline the flow of imaging studies and optimize load-balancing.[[{"type":"media","view_mode":"media_crop","fid":"45562","attributes":{"alt":"radiogenomics","class":"media-image media-image-right","id":"media_crop_8786454422882","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"5219","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":"width: 200px; height: 199px; border-width: 0px; border-style: solid; margin: 1px; float: right;","title":"©Lonely/Shutterstock.com","typeof":"foaf:Image"}}]]

RSNA 2015 Vendor Highlights:  Medicalis, which is ahead of the curve, is finding its sweet spot in large distributed health systems that conduct more than 1 million annual imaging procedures. Meanwhile, traction is accelerating across various segments of the market on the part of few other vendors including McKesson Conserus, Intelerad, Primordial, and Clario. The latest entries into this field of Mach7 Technologies and Infinitt North America, and the ‘DICOM plumbing’ provided for this purpose by Laurel Bridge is worthy of honorary mentions, too.

• Next-generation image streaming technology is bound to supplant traditional PACS image servers and conventional server-based and Web-based clients. The current situation is reminiscent of a decade ago, when the then cutting-edge wavelet technology of Stentor was disrupting the marketplace before it became the PACS giant that Philips iSite is today. The recent advances in PACS streaming technology take advantage of the latest developments in server-side computing and ultra-thin clients.

RSNA 2015 Vendor Highlights: Although the subset of enterprise viewers that have gained FDA approval for primary image interpretations continues to grow, these viewers are still primarily used for secondary use cases. Joe Marion, Principal at Healthcare Integration Strategies, has short-listed the ones he believes have the right streaming technology to be a viable PACS replacement solution: Visage Imaging (with a three-year head-start), Fujifilm’s latest Synapse PACS, and Viztek (now part of Konica Minolta).

• Cloud-based imaging informatics technologies have become robust and reliable solutions, not only in their private and hybrid configurations, but in their public ones as well. CIOs have come a long way in accepting the concept by not rejecting it outright, and PACS vendors have become much more comfortable, if not proactive, in proposing cloud storage options. The extent and the pace at which the use cases for cloud-based imaging have evolved over the last five years is quite amazing, from merely providing back-end archiving functions, to much higher-level use cases such as image exchange, remote 3D imaging services, teleradiology, PACS-as-a-service, analytics as-a-service, and other software-as-a-service (SaaS) applications.

RSNA 2015 Vendor Highlights: We, at Frost & Sullivan, are impressed with Dell Healthcare’s continuing success in transforming itself from the cloud-based image archive it was 10 years ago, into the full-fledged cloud platform that it has become today, where more third-party application vendors are plugging-in to create a best-of-breed service bus. GE Healthcare’s U.S. launch of the Health Cloud is exciting as well. It is touted as the first large-scale, purpose-built cloud for health care, which leverages GE’s Predix platform to emulate GE’s industrial Internet efforts in other industries such as in aviation.

These various imaging informatics advances collectively point in one direction-they give access to a unified view of larger pools of images, and the means to leverage these datasets in a way that is more effective and less hardware-intensive than in the past. The logical next step is to unleash their full potential by applying analytics and machine learning to these medical images.

Enter Radiomics at the Crossroads of Cognitive Computing, Big Data, and the Cloud
Going back to the ‘WOW!’ effect, RSNA 2015 did not disappoint, particularly at the IBM/Merge Healthcare booth. Surprisingly, only a few months into the acquisition, the combined entity has already gone a long way in applying IBM Watson’s cognitive computing and machine-learning algorithms to imaging. The live booth demonstration of the works-in-progress showed how deep image analysis using image quantification, segmentation, classification, pattern recognition, and characterization can be used to propose statistically significant guidance on what the most-likely diagnoses would be, based on comparisons with a large cohort of other patients’ images and associated reports.

This type of application can be dubbed as next-generation clinical decision support (CDS), so as to differentiate it from traditional computer-aided detection applications that do not leverage the big datasets sitting in the cloud in real time. CDS currently stands out as the most promising use case across many applications, whether it is deciding the next-best-step in navigating the care pathway, deciding on the most likely diagnosis, appropriate ordering of imaging studies, stratifying patient risk or triaging patients for population health management.

The Way Forward Towards Precision Medicine with Major Implications on Personalized Medicine
A whole new competitive field is forming around this emerging and hugely promising new opportunity for medical imaging. It holds, by many accounts, the key to move the field decisively into the era of value-based health care.

The top-ten list of vendors and initiatives we are tracking with the highest degree of interest in this field are IBM/Merge Healthcare, Enlitic, RadLogics, HealthMyne, vRad/MetaMind, Zebra Medical Vision, Mindshare, Butterfly Networks, Siemens teamplay, and Fujifilm Case Match. While most of these companies are starting with specific imaging modalities in limited disease areas or are in a piloting phase, they are making waves and moving fast. The race is on!

The upcoming HIMSS 2016 should be interesting in tracking these trends, especially in gauging where imaging stands within the greater health care analytics movement, and determining how radiomics and genomics may combine to create the first real-life applications of radiogenomics.

Related Videos
Can Diffusion Microstructural Imaging Provide Insights into Long Covid Beyond Conventional MRI?
Assessing the Impact of Radiology Workforce Shortages in Rural Communities
Emerging MRI and PET Research Reveals Link Between Visceral Abdominal Fat and Early Signs of Alzheimer’s Disease
Reimbursement Challenges in Radiology: An Interview with Richard Heller, MD
Nina Kottler, MD, MS
Related Content
© 2024 MJH Life Sciences

All rights reserved.