This Insight is part of your subscription to Signify Premium Insights – Medical Imaging. This content is only available to individuals with an active account for this paid-for service and is the copyright of Signify Research. Content cannot be shared or distributed to non-subscribers or other third parties without express written consent from Signify Research. To view other recent Premium Insights that are part of the service please click here.
Co-written by Amy Thompson, Dr Sanjay Parekh & Steve Holloway
A new year has arrived. Despite the numerous new releases and innovations throughout 2021, and vendors laying foundations for longer-term technological shifts, vendors enter 2022 under similar circumstances to 2021. A new variant of Covid-19 is sweeping the world, causing disruption to both healthcare services as well as the wider economy. Providers are looking for efficiencies and must find ways to address the enormous and ever-present backlog of patients, while vendors must raise their eyes and continue to deliver on longer term strategies, without neglecting providers’ needs of the moment. Despite this uncertainty, however, there are several key trends and developments we can expect to see in the coming year.
Vendor focus will intensify on structured data aggregation
Vendors and providers alike are increasingly understanding the potential that clinical data offers healthcare. This is true not only in a singular department such as radiology, where the use of data has, in the past, primarily focused, but throughout the wider clinical workflow, including in preclinical and clinical research, companion diagnostics and precision medicine.
To realise this potential there must first be progress in several other areas. One of the most pressing challenges will be getting standardised and machine-readable data from the radiologist reports and automating the population of the findings into the clinical workflow in the first place. Structured reporting tools will help facilitate this inclusion of standardised diagnostic data into the clinical workflow. Over the course of the coming year, vendors will continue to develop structured reporting technology, either in house, or through partnerships that help lay the foundation for a greater role of data in diagnosis. Further, data management platforms will continue to technically evolve, focusing not only on image management, but broader structured “holistic” data management of images, annotations, reports and clinical documentation.
Federated learning also looks set to be increasingly adopted by vendors. This will enable AI developers to create generalisable AI solutions without the need to host third-party data. As such, federated learning will enable AI developers to leverage healthcare providers’ local data to train their algorithms without that data leaving hospital infrastructure. This means that algorithms can be trained on larger and more diverse datasets that are more representative of local case mix, ensuring that the algorithms have a high sensitivity and specificity.
Thirdly, a host of new partnerships will be made between imaging vendors, healthcare providers and the pre-clinical sector (clinical trials, CRO’s, pharma). Demand for Real World Evidence (RWE) to support a new era of drug-development and companion diagnostics will present new commercial opportunities to leverage the vast datasets held by imaging vendors and providers alike.
Pathology will play a greater role in medical imaging IT
The longer-term direction of medical imaging IT systems to fully fledged enterprise imaging solutions is well understood, but so far, the number of solutions that have expanded beyond radiology and into other departments is minimal; the combination of radiology and cardiology has been the primary example from vendors when displaying multi-ology functionality. This will begin to change in 2022, with digital pathology becoming a focus area for many vendors and where significant progress will be made.
Digital Pathology is being adopted in three ways: primary diagnosis, supporting pathologists’ primary reading and reporting; secondary use, including clinical consult, tumour boards and medical education; and preclinical use, in research, clinical trials and drug development. Adoption has been ramping up significantly in preclinical use. However, the limited deployment of primary use systems pre-Covid limited the scope for secondary adoption and integration into enterprise imaging.
The advent of the global Covid-19 pandemic has dramatically shifted provider focus. There has been much greater investment in diagnostics and the digitalisation of pathology with the inefficiencies of lab processes highlighted during the pandemic. Furthermore, loosening regulation for primary use in the USA has opened-up the largest market globally, while VC/PE investment in digital pathology reached record levels in 2021.
For imaging vendors, digital pathology will increasingly become a key topic and differentiator in enterprise imaging deals in 2022. For the majority, offering primary use applications will only be tenable via partnership, so we expect to see more deals announced between imaging IT vendors and specialist digital pathology vendors in 2022, following in the wake of announcements such as that of Fujifilm and Inspirata.
Furthermore, we expect to see more focus on bespoke workflow tools that support the use of pathology data and reports in radiology, as well as broader multidisciplinary workflows (e.g., tumour boards and alike). Fundamentally, most healthcare providers in mature markets are still early in terms of primary and secondary use adoption of digital pathology. However, most are looking to leverage existing and future investment in enterprise imaging where possible to deploy an integrated offering that can support access to digital pathology and integrate a primary diagnostic offering. Therefore, vendors must be prepared with a clear roadmap for how to address this need for enterprise imaging capability to support secondary use, while also supporting primary use system integration. With many prominent EI deals due for renewal or tendering in 2022, competency around digital pathology could be a dealmaker or a deal-breaker.
AI product categories will blur as Imaging IT vendors increasingly control channel
AI is still a very young market within medical imaging. As such, the technology itself as well as its application and use are evolving quickly. Leading AI tools have already progressed from being single point solutions which assess a medical image to identify a single radiological finding, into more complete and more mature solutions which offer greater value to providers. At present these more sophisticated solutions may be categorised into one of three categories, which exist alongside advanced point solutions, such as FFRCT:
– Comprehensive solutions that can identify many findings on an image and make a more meaningful difference to a radiologist’s workflow than single point solutions, whose clinical value is often limited. Additionally, they may pick up secondary or incidental findings that the radiologist may not even be actively looking for, avoiding missed diagnoses.
– End-to-end solutions, which offer care coordination and solve problems along the length of the clinical workflow, adding value beyond the diagnostic slice, for a given clinical condition such as stroke care.
– AI body area suites or workflow packages, typically provided by imaging IT vendors, which curate multiple native and third-party AI tools and other capabilities
These are currently well-defined categories, but over 2022, the distinctions between these groups will become less clear. Comprehensive solutions will begin addressing providers’ requirements outside of the diagnostic portion and expand into care coordination and clinical decision making. Third party platforms and marketplaces will combine more specialist tools from vendors into suites that are able to address many different aspects of a high value use case, by combining several single-use solutions. Workflow packages will expand beyond single use cases, creating greater value solutions for radiologist.
Beyond this evolution of the solutions themselves, however, 2022 will also see imaging IT vendors taking a more central role in the deployment of AI and enabling its broader use. This will see these vendors expand comprehensive solutions to take on an end-to-end approach. In addition to image analysis capabilities, imaging IT vendors could also leverage worklist triage, structured reporting, advanced visualisation and other workflow components in more cohesive packages creating further value for the radiologist.
AI tools will not be sold on the basis of their technology
Until recently, medical imaging AI solutions have advertised themselves based on their use of machine learning technology and accuracy. However, in a transition that will accelerate and become more prevalent in 2022, the focus will be instead on the clinical advantages of the products. AI developers will, instead of selling a technology, highlight their solution’s ability to solve problems for providers.
Providers are becoming more knowledgeable about AI, and the number of vendors offering AI solutions has increased dramatically over recent years, so highlighting a solution’s machine learning credentials alone is no longer enough to differentiate a product in the market. Instead, vendors are having to highlight the overall value they can bring to a provider. This is a challenging task in many cases, especially as many providers currently do a poor job of measuring clinical and diagnostic outcomes accurately.
This has been seen already in stroke care, for example, where some vendors have moved away from touting the technical performance of their stroke detection algorithm, and instead positioned themselves as stroke care vendors by creating a care coordination platform, emphasising their capability across the entire stroke care pathway. Crucially, the importance of the technical specifics of these stroke care vendors’ algorithms have been displaced by a metric that providers are much more interested in; a solution’s ability to deliver the best outcomes for patients.
This trend will gather pace over 2022 and will be prevalent across acute (e.g., pulmonary embolism) and chronic (e.g., oncology) use cases. While algorithm developers will be keen to show that they are able to integrate into medical imaging departments with more thorough solutions, other companies such as marketplace and platform providers will also look to better meet providers’ needs. This need to offer solutions that solve vendors’ problems mean that partnerships continue to proliferate as AI vendors are too small or too specialist to offer the broader clinical capability to providers look to partner to bolster their value proposition.
Operational workflow’s importance will only grow and increasingly influence business models
There were several releases of operational workflow tools in 2021 as imaging IT vendors sought solutions that could improve the efficiency of their customers. These vendors have focused on developing solutions which grant providers better oversight of the operation of their medical imaging departments, their radiologists, and their fleets of medical imaging equipment. The continuing development of these solutions will allow these providers to attend to more patients and more precisely manage their patient’s care pathways.
The development and inclusion of these tools are, in the longer term, also set to give providers better oversight of their departments and generate data about their operation, which will lead to better planning capability, and improve their ability to strategise. This will become particularly important as acute providers utilise teleradiology or outpatient imaging centres for additional reading, or as provider networks grow through M&A activity and visibility over multiple sites is required.
To make the most of this demand for operational tools from providers, vendors will focus on ensuring their solutions are, for the most part at least, vendor agnostic. This will enable these vendors to sell their solutions into any provider, regardless of which other imaging IT systems they use, and grant the vendor an entry point into providers served by competitors. However, the incumbent imaging IT vendor is still best placed to offer providers the richest feature set, making it the most likely to be chosen by providers.
The importance of such solutions will be highlighted in 2022, but different vendors will take different approaches. Vendors which specialise in imaging IT and are targeting the development of enterprise imaging systems will focus on worklists, triage and departmental or operational analytics, while the vendors which also offer hardware will ensure that the use of their modalities, particularly those that are more time consuming such as MR and CT, can be optimised and departments can maximise the use of their resources. These vendors will look to hone functionality such as virtual image acquisition, automatic patient position and protocol efficiency.
The intensified focus on workflow will also present more opportunities for layering on professional services to operational workflow deals, placing vendor specialists within a provider organisation to support improving operational efficiency and workflow evolution. This will be the first step towards tighter vendor-provider relationships and the advent of contracting and business models based on specific performance indicators and operational targets. While a significant risk for the vendor, providers are already coming round to “performance” focused contracting, thereby ensuring a level of service and operational competency while de-risking capital investment. Therefore, providing analytics and workflow tools that identify inefficiency within radiology service-lines will no longer be enough, providers will expect their vendor partners to help solve these problems and support more efficiency care delivery.
About Signify Premium Insights
This Insight is part of your subscription to Signify Premium Insights – Medical Imaging. This content is only available to individuals with an active account for this paid-for service and is the copyright of Signify Research. Content cannot be shared or distributed to non-subscribers or other third parties without express written consent from Signify Research. To view other recent Premium Insights that are part of the service please click here