Two takeaways from RSNA that will shape enterprise imaging
As with every year, RSNA provided a glimpse into the future of radiology. Undoubtedly, AI was the most discussed and debated topic, though for imaging informatics there were some other less obvious trends on show that will have a significant bearing on the future of the market. Here’s my top two takeaways from the show:
One of the most striking observations from wandering the exhibit halls was the almost non-existence of “PACS” and “RIS” in any marketing—“enterprise imaging” has very much taken on “AI” and “VNA” for the most over-used and often under-deserved product title. Does this signal that PACS and RIS has finally been confined to the annals of informatics past? Or is there more to the sudden explosion of enterprise imaging products than meets the eye?
Put simply, PACS is still very much alive and kicking and still forms the core of most imaging informatics platforms that were on show. However, because of increasing integration and crossover with allied products such as advanced visualization, universal viewers, VNAs and workflow tools, the definitions between distinct product groups have become blurred.
In the North American market as well as other mature international markets, growing demand for enterprise products in the wake of enterprise electronic medical records (EMR) adoption has driven a frenzy for “enterprise” products in imaging informatics. Yet this is not the case when you look at the market data–86 percent of revenues spent on imaging IT worldwide in 2017 were for standalone software modules (such as PACS or RIS), with “enterprise” software only accounting for the remaining 14 percent. Stronger growth is expected for enterprise products in the coming years, but the imbalance in marketing on show is far removed from the reality of use today.
Most solutions exhibited at the show described as “enterprise imaging” are little more than multi-site PACS systems, or an evolution of multi-site PACS with additional features or partnerships with third parties to support remote viewing (universal viewers) or centralized archiving (usually a DICOM archive or VNA). Very few have the technical maturity to support the reading, management and storage of non-radiology (non-DICOM) images and unstructured or complex clinical content from other departments.
This is perhaps exemplified by the level of maturity in integration with different clinical departments of many of these solutions. Take for example the case of breast imaging; digital breast tomosynthesis (DBT) and automated breast ultrasound (ABUS) analysis now make up a significant portion of the adaptive breast screening pathway and have been in regular use for some time, yet still many products marketed as “enterprise imaging” have no capability to manage or integrate native images from these modalities.
Nuclear medicine is another common source of imaging that has been overlooked from an integration perspective; the same too can be said of endoscopy and surgical video, point of care ultrasound and digital pathology. Until integration maturity improves to support imaging in these departments, vendors should perhaps stick to a clearer and more consistent name for many of these products, such as “enterprise radiology” rather than “enterprise imaging.”
That said, there were positive discussions at many booths around requests from customers to bring a broader set of clinical data into the diagnostic reading environment. Semantic search tools to pull in relevant patient data from the EMR, dependent on the type of scan displayed has already been demonstrated by several vendors utilizing AI, with some even enabling access to oncology and pathology reporting within the imaging IT platform, in order to provide a richer diagnostic context for reporting radiologists.
This is certainly a positive shift in terms of patient care, though it has a long way to go in terms of implementation. Many radiologists today struggle to handle the vast amount of information provided by today’s advanced imaging modalities and are under pressure to read quicker, so information overload is a major concern. Interoperability between the EMR and diagnostic departments such as radiology also remains a barrier—loosening of data protectionism from EMR vendors has certainly helped of late, but there is still a long way to go.
Combined, these factors all suggest that customers and vendors alike should be wary of the broad term of “enterprise imaging,” especially when loosely applied to any product that is deployed in a multi-site environment. Instead, the industry needs to better define the maturity of solutions based on their technical capability to both scale and ability to manage imaging and associated clinical content.
Very few health providers or vendors have implemented “true” enterprise imaging in practice today, even if that is their intended direction of long-term development. Instead, “enterprise radiology” would be more apt to depict the capability of solutions in use today, as a stepping stone towards more complex enterprise imaging. If not, many may find that their choice of “enterprise imaging” solutions is akin to the emperor’s new clothes for the Danish legend; little more than an enterprise radiology solution in an invisible “enterprise imaging” suit.
Health providers are under pressure to deliver high-quality services to more people at lower cost, meaning radiology is increasingly under the microscope as an area for providers to make operational efficiency gains. However, many business intelligence and operational workflow product developments for were not front-and-centre of booths in the exhibition hall – instead they were embedded deep in software version updates of existing imaging IT platforms or positioned as new “add-on” workflow or operational modules. We can group many of these additional capabilities into the following broad categories:
Radiologist workflow tools: auto-triage, adaptive menus, anatomically adaptive hanging protocols, peer-review and collaboration tools
Operational tools: Case-load balancing, enterprise worklists, smart scheduling, order entry integration, modality fleet management, modality protocol editing, staff resourcing
Business intelligence and practice management: department performance dashboarding, staff and clinical audit, regulatory compliance, finance and billing support, quality and outcome measures
While these tools may not be exclusively used by radiologists, they will all have a significant bearing on patient safety, operational cost and care efficiency. Care provision today is increasingly focused on care quality and efficiency, so the ability of software and services to support ongoing improvements in how radiology departments are run has become as important as diagnostic functionality. Vendors have realized this is an area that so far has been underserved in imaging and offers a significant opportunity to tap into the operational budgets of imaging services at health providers, as opposed to just their IT and informatics budgets.
There were numerous examples on show of improvements in updated platforms. In terms of radiologist workflow tools, multimedia structured reporting, adaptive toolsets and displays (such as anatomical recognition and adaptive menus) and increasingly “smart” worklists are becoming more common. In the business intelligence and practice management segment, performance analytics and dashboarding has becoming increasingly available and specific for the radiology department.
Most progress so far has centralised around a few key areas; staff audit and training, especially related to dose awareness and compliance, are now increasingly integrated and expected from radiology administrators; integration to financial and revenue cycle management has also become more prominent. Other novel features and tools are being released too, such as analytics and dashboarding from basic patient admission information, in order to find patterns or reasons for missed appointments and potentially missed revenue. These are also being paired with smart scheduling for appointments and follow-up to minimize downtime and reduce patient “no-shows.”
In terms of operational workflow, the push from vendors towards enterprise platforms (be they enterprise radiology or enterprise imaging) has also made case-load balancing of enterprise worklists and centralized scheduling more commonly available. Perhaps above all, the leading big-iron modality vendors were all exhibiting some form of modality fleet management.
Operational diagnostics and service reminders to support efficiency for CT and MRI have been around for some time, but newer features such as centralized protocol management (as seen in Philips’ Healthcare PerformanceBridge module) enable radiology administrators and expert technicians to manage and monitor multiple big-iron modalities from a single central location.
Siemens Healthineers was perhaps the most forward in this thinking in terms of fleet management, taking it one step further with their “syngo Virtual Cockpit,” a software module that allows a remote expert technician to view (via video feed), assist or manage scan preparation, protocol management and offer real-time support to on-site staff. Initially a concept designed to support advanced scanning in remote locations or emerging markets with limited expert technicians, it is feasible that vendors will also look to offer their own experts to support modality monitoring as a service soon.