Market Insight

In 2018, Analytics and AI in Healthcare Look to Take Center Stage

January 04, 2018

Jason dePreaux Jason dePreaux Chief of Research, Industrial Technology Solutions

Want to learn more?
Have an expert contact you.

The 103rd Scientific Assembly and Annual Meeting of the Radiological Society of North America (RSNA) brought together over 50,000 attendees at Chicago’s McCormick Place from November 26th to December 1st, 2017. Companies of all sizes brought their A game to the exhibit floor with leading companies showcasing gargantuan booths filled with the latest and greatest medical imaging technologies.

Despite the lingering lack of consensus as to its true meaning, the permeation Enterprise Imaging was evident throughout the conference. Dominant medical imaging companies such as Agfa Healthcare, Carestream, GE Healthcare, Hologic, Philips and Siemens Healthineers revealed their increased footprints and intention to proliferate across different “ologies” in healthcare through use of interoperable standards.

The propellants of increasing demand for enterprise imaging systems are: the desire for greater functionalities across disparate clinical areas, and the need for enhanced visibility into the operations of the entire enterprise. As healthcare providers grow and become more value-based, the ability to collect standardized performance measurements is crucial to benchmark metrics in all clinical domains. For these reasons, successful deployments of enterprise imaging systems require the support of robust analytics platforms.  

IHS Markit predicts that on-demand, real time insights into operations and workflow optimization will dominate conversations in medical imaging in the near future. Analytics, coupled with increasing use of artificial intelligence (AI) are two key forces to watch in 2018.


Augmentation and Optimization: the two promises of analytics and AI

In addition to providing visibility, a well-designed analytics platform is an inherent decision support tool that conducts root cause analyses and identifies bottlenecks. It reveals actionable insights into the potential causes of inefficiencies and ways in which issues can be resolved.  A few increasingly recognized features of robust analytics platforms are: 

  • Vendor agnosticism—the prime target users of analytics are large, growing integrated delivery networks (IDNs) which means that the ability to connect with multiple vendors across the enterprise is key to analytics success.
  • Auto-generation of highly configurable dashboards with nuanced understanding of (or ability to adjust to) local clinical contexts across various members of an IDN.
  • Ability to benchmark and compare performance both intra-and inter-enterprise of all sizes and geographies.

A few examples of companies that were showcasing analytics platforms at RSNA include:

  • Agfa Healthcare: Business Intelligence and Workflow Engine
  • NTT Data: Unified Clinical Analytics and Management Platform
  • Philips: Performance Bridge
  • Siemens Healthineers: Teamplay and Screening Workflow Navigator
  • Watson Health Imaging: Merge Dashboards

As medical imaging hardware becomes more advanced and analytics are used to better understand operations, AI is introduced to this realm as a technology which augments radiologists’ workflows and automate mundane tasks providers face. Through deep learning, the early use cases of AI in medical imaging is to reduce the amount of information providers need to evaluate in this multidimensional space. It was evident throughout the show that companies took careful considerations when touting about the benefits of AI, primarily highlighting its assistive powers rather than defining it as a diagnostic tool. The newfound focus on workflow optimization and operational efficiencies also provide a point of entry for larger technology companies such as Google and NVIDIA to enter the healthcare arena.

While promising, AI use in healthcare remains constrained by the lack of unified data architecture. Particularly in radiology where structured reporting is more immature compared to other discipline such as cardiology, use of AI will remain limited until the issue of unstructured data is addressed.

At RSNA2017, some of the notable announcements regarding AI use in medical imaging were:

  • Agfa Healthcare showcased its ecosystem with integrated Augmented Intelligence:
    • The work-in-progress reveals the integration of Agfa Healthcare’s Enterprise Imaging solution with products offered by IBM Watson Health, RadLogics and Mindshare Medical.
  • Change Healthcare announced its partnership with Zebra Medical Vision and Google Cloud to create tools using AI:
    • The partnerships will focus on ways in which AI can be introduced to better connect and automate clinical workflows.
  • GE Healthcare announced its partnership with NVIDIA:
    • The 10-year partnership announcement will bring AI to GE Healthcare’s 500,000 imaging devices globally and accelerate the speed at which healthcare data can be processed.
  • NTT Data announced its partnership with
    • In addition to the incorporation of AI as automation tools in the delivery of clinical care, this partnership includes the ability for NTT Data Services clients to share revenues for use of anonymized studies in data and algorithms validation.
  • NVIDIA, in addition to its partnership with GE, sees increasing opportunities for its Graphics Processor Units (GPUs) in AI healthcare applications due to their power efficiency. At the large end they are targeting Volta GPUs for imaging algorithms in a more traditional data center environment. On the embedded front, their Jetson platform can power portable medical devices.
  • Other AI companies that offer more niche-specific algorithms were:
    • Aidence, Arterys, Blackford, 4Quant, aidoc, ContextFlow, Combinostics, CureMetrix, DiA Analysis,  iCAD, icometrix, imbio, Infervision, Kheiron Medical Technologies, LPixel, Lunit, Quantib, Quibim,, RadLogics, Riverain Technologies, and VUNO


Our Take

One of the main challenges in commercializing AI platforms in the budget-constrained healthcare industry will be the ability for companies to convincingly articulate ROIs in ways that justify the investment. With this obstacle, we expect to see greater risk-sharing agreements with long-term strategic collaborations that allow providers and vendors to work together in tailoring solutions that best address the needs of individual providers and networks. Partnerships with a diverse mix of institutions will be key to establishing the successful use cases for analytics and AI—although tangible results will take a lot of time and resource commitments with uncertain expected outcomes.

As the realm of enterprise imaging continues to evolve, analytics and AI are poised to become two key differentiators that will influence purchasing decisions. Considering the growing commoditization of PACS within mature markets, innovations in these technologies can further enhance replacement incentives and unlock future growth of the $2.8 billion market for Radiology IT.

IHS Markit will continue to follow these trends as part of our broader decade-long coverage of the healthcare IT industry, which comprises of syndicated reports on: Medical Enterprise Data Storage (VNA and Image Exchange), EHRs, and Radiology & Cardiology IT (RPACS, RIS, CV-PACS, CVIS). For more information, please contact Nile Kazimer at [email protected]

Share facebook Twitter Google Plus Linked In Add This Contact Us