Share article:
Sidetracked
With healthcare IT becoming increasingly enterprise across departments and sites, Advanced Visualisation (AV) as well as PACS, is commonly a strategic obstacle due to its often standalone workstation-based nature. The wish for diagnostic viewing integrated into the healthcare enterprise with multiple access points, has given new hope for the more lightweight Universal Viewers (UV), and indeed increased demand for UV with diagnostic tools. The UV market is still small compared to the more mature AV market, but after years of relying on superior diagnostic features alone, AV vendors are now realising AV needs to change in order to remain as the primary diagnostic tool. Just like PACS, AV cannot continue to be a disconnected island within hospital informatics. And what do you do when you realise you are being sidetracked? You integrate. RSNA this year testified to this realisation with AV showcases and announcements focusing on AV integrating more tightly in the workflow pre- and post-diagnosis, as well as new developments with AI assisted image analysis.
Upstream Integration
As RSNA President Dr Rao noted in her opening speech about radiology 25 years from now, radiologists will start the day as they do today by clicking through the list of cases. However, AI will have helped prioritising the worklist based on pre-analysis and automatic measurements, and already have pre-diagnosed based on similar cases and supporting clinical information. We can therefore expect worklists to present the results from dynamic imaging IT hubs pulling in information from different clinical sources such as EMR, genomic sequencing, family history etc in order to perform an initial imaging and context analysis, prioritising the cases and presenting them to the radiologist for further analysis using AV if applicable.
Over time, we can expect the current PACS, UV and AV viewers to fuse into one diagnostic viewing platform with accessibility throughout the enterprise, and these viewers will be tightly integrated with the worklist, partly mediated by the underlying AI platform providing pre-analysis for the worklist and visibility in the viewer. The worklist will therefore grow dramatically in importance from the early PACS/RIS worklists and become a cornerstone in clinical enterprise IT. Intelerad’s acquisition of the zero-footprint worklist company Clario Medical, announced during RSNA, can therefore be seen as an attempt to build on Intelerad’s recent success as an enterprise Imaging vendor and investing that not only in a new tool and local presence in the US, but in a strategy of integrating their solutions deeper in the future radiologist workflow. This follows an ongoing trend of worklist vendors being acquired, or Imaging IT vendors developing own workflow tool modules. Last year Siemens Healthineers completed the Medicalis acquisition, Change Healthcare developed their Workflow Intelligence dynamic worklist, and Mach7 offered their Universal Worklist, amongst others.
Lateral Integration
AI is still very much in the hype phase where everything is possible, and everyone wants to be associated with it. However, there is starting to emerge more pragmatic discussions like “how do we integrate it into the workflow”, and “how should radiologists approve, deny or give feedback to the AI generated results”, which indicates the technology is maturing and establishing itself in the industry. For years to come AI will most likely be involved in case prioritisation or for routine screening, possibly replacing some of the simpler tasks, but it will not replace the more complex image analysis AV is typically being used for today. AI and AV will therefore need to coexist and complement each other as a natural part of the same workflow.
Most early standalone AI products available today require uploading of images to the cloud for analysis and return a complete quantification report. However, there were also examples on show of partner AI algorithms from vendors such as Koios, Qure.ai, iCAD and HeartFlow being tightly integrated into AV platforms and diagnostic viewers such as Visage7 (Visage Imaging), ResolutionMD (PureWeb), AW (GE Healthcare) and syngo.via (Siemens Healthineers) working side-by-side with the AV clinical tools. EnvoyAI, following its acquisition by Terarecon, is going for a slightly different approach, using the Northstar AI Explorer for visualisation of AI results from the EnvoyAI marketplace and integrating the Northstar viewer with either third party PACS or Terarecon’s own iNtuition AV solution. GE Healthcare announced their new brand name Edison which will include many of their existing clinical AV tools, native AI applications, partnership apps and future imaging analysis and predictive and operational analytics applications, representing an approach similar to other big iron vendors in keeping a centralised AI platform, which then integrates with its AV solution.
Downstream Integration
Many clinical apps on the market have their own reporting tools embedded as part of their solution, however many radiologists use their PACS reporting tools or third-party reporting tools across the different clinical areas for their reporting. As there is much variation in reporting guidelines between countries and clinical areas, some local vendors with structured reporting solutions thrive. However, the market is largely consolidated around a few major players, with Nuance the dominant force in the US market. Philips announced last year a partnership with Nuance for structured reporting in their AI platform Illumeo, and Siemens Healthineers showcased at this year’s RSNA how tight integration between Nuance Powerscribe and Siemens syngo.via could create a direct and semi-automated reporting link between the AV platform and the downstream reporting procedure, auto-populating quantifications and observations in the common reporting environment. Additionally, Carestream showcased it’s latest multimedia reporting features in Vue Reporting. This suggests that leading vendors in the AV and broader imaging IT industry are becoming increasingly serious about enabling semi-autonomous, structured reporting in radiology.
Tighter Workflow Integration Enables
The focus on workflow integration and AI were not randomly picked topics to entertain the attendees at RSNA; rather it’s part of the overall strategic thinking of AV and medical imaging vendors that they need to integrate AV tightly into the enterprise structures and have continued focus on creating value for their customers, by enabling them to save money on operations, do more diagnoses with fewer resources and improve the standard of care. Streamlining the workflow and incorporating AI for image analysis into AV helps the vendors achieve this goal, while also moving AV from a disconnected island, to a central role in the future of diagnosis.