This Insight is part of your subscription to Signify Premium Insights – Medical Imaging. The content is only available to companies that have subscribed to this paid-for service. To view other recent Premium Insights that are part of the service please click here.
Last month saw the release of Signify Research’s Competitor Ecosystem topical report, one of the deliverables of the AI in Medical Imaging Market Intelligence Service. With the medical imaging AI market evolving, there are several trends that are standing out and impacting the complexion of the market. Such maturation has left vendors with striking opportunities in the market, as well as some significant challenges.
The Signify View
One of the most dramatic indicators of the development of the market is the ability of some vendors to raise investment to levels almost unimaginable just several years ago. Tellingly, the nature of investment is also changing. Previously, the focus of private investors centered around smaller, earlier-stage funding rounds for young start-ups. More recently, however, the emphasis for investors has shifted toward supporting larger, better-established vendors with more sizable later-stage rounds.
This trend has given rise to a select group of vendors. Companies which, by one metric at least are market leaders, having raised more than $100m in investor funding.
While the number of vendors in this rarefied air is increasing, even among these investor favourites, exceptional performers are starting to emerge, with several vendors having individual funding rounds of more than $100m. Rounds that, individually, dwarf the total amount of funding that more than 95 percent of any other companies have been able to raise in total. This level of investment, propelling this select group of vendors forward highlights the confidence investors have in these companies as businesses, not just technological innovators. Such sizable later stage investment shows that investors realise that select AI firms not only have compelling technology, but also a robust product portfolio and represent a strong value proposition. There are exceptions to this trend, with some, primarily Chinese, vendors benefitting from government incentives and support, but in most cases, such impressive levels of investment demonstrate vendors that have been able to turn technology into business.
It’s Getting Later
In addition to consolidating the position of market leaders, this trend of private investment increasingly focusing on later stage funding rounds and already established vendors means that smaller vendors will find it ever more difficult to secure funding. Over time this will leave them facing difficult and potentially desperate decisions. In a developing tale of the haves and the have-nots, smaller vendors which have yet to gain commercial traction or develop sophisticated solutions with their core technologies will face shrinking funding runways. As this happens, they will cease to be able to afford the costly product development and expensive clinical validation studies that will enable them to grow and rally. Over time, these shrinking runways will lead to dramatic market consolidation as vendors become more open to acquisitions, pivots, or, if necessary, dissolution.
This consolidation around the existent leaders is happening in other ways too. While demonstrating their commercial potential to investors, financially well-supported vendors have also shown their market leadership calibre in another way. Many of these AI developers are increasingly able to tout not just regulatory clearances, a milestone which essentially demonstrates that products work safely and as intended, but also reimbursement. This is significant. One of the most difficult barriers for vendors to overcome is that of convincing providers to pay for their solutions. While reimbursement does not necessarily make providers money, it offsets or at least mitigates, to some extent, the cost providers must pay to take advantage of AI solutions.
The awarding of reimbursement to solutions, combined with the capabilities of the tools themselves, will help motivate providers to adopt and help AI become an increasingly mainstream tool. This will help consolidate certain providers’ positions as market leaders, and set those vendors that have failed to innovate, even those that started strongly, further back in their paths to reach market leadership positions. This could be particularly true as market leading vendors look to expand the breadth of their portfolios, potentially encroaching on markets targeted by smaller competitors.
The Guiding Hand
Reimbursement also has the potential to be transformative in other ways. In addition to mitigating the cost barrier stymieing adoption of medical imaging AI at providers, through reimbursement, regulators can also guide the markets they oversee, encouraging and essentially subsidising development in certain directions. This has recently been apparent in the US, where reimbursement has been awarded to solutions such as Cleerly’s cardiac plaque detection algorithm and Optellum’s tissue characterisation algorithm. While both tools are very different and have very different clinical uses, the fact that both offer advantages in a broader, population health context, rather than simply offering advantages in very specific contexts with limited downstream impact will no doubt have helped solidify their case for reimbursement.
Further by offering reimbursement for AI solutions that also offer advantages in a broader population health context, vendors will be encouraged to address this consideration as they continue to develop their solutions. There are similar motivations with regards to other tools, with, for example, several solutions which have received reimbursement reshaping the established diagnostic pathway, allowing a shorter time to treatment, in the case of stroke algorithms, or, in some cases, eliminating the requirement for invasive diagnostic procedures, as is the case for FFR-CT.
Reimbursement is helping to overcome the cost barrier that is holding back the adoption of medical imaging AI, but there are also other challenges slowing the pace of the technology’s uptake. One of these can be the difficulty of deploying AI into the clinical workflow. AI platforms have become increasingly common as vendors look to solve the last-mile challenges of deployment, integration, and orchestration.
As these platforms continue on their way to ubiquity, they are, like the solutions they deliver, also becoming increasingly sophisticated. Many platforms initially served the relatively straightforward purpose of becoming a means to host applications and making them easily accessible to providers. Latterly, however, platforms are serving more complex services. One trend, for example, has seen algorithm developers themselves begin to offer commercial platforms.
This is a logical progression, with algorithm developers having to essentially offer platforms as their range of native tools grew and they needed an efficient way to be able to deploy them all into providers’ workflows. Some vendors have expanded this functionality commercially, hosting third-party algorithms alongside their own natively developed solutions. By bolstering their platforms in such a way these algorithm developers can further improve the clinical utility of their offerings, using third-party applications to supplement their own natively-developed capability, and in doing so offer curated packages and workflow suites tailored to particular clinical workflows.
Instead of simply deploying AI into workflows, these more sophisticated, better curated platforms can orchestrate algorithms, to not only deliver capability, but ensure it can be effectively utilised. Over time, in many cases these platforms, whether natively developed by informatics vendors, specialist platform providers, algorithm developers or even modality vendors, will replace the direct integrations that characterised the early days of medical imaging AI adoption. In this way platform providers will harbour increasing sway over the medical imaging AI market.
This trend could begin giving large imaging IT vendors reason to start making acquisitions, beyond simply acquiring specific AI capabilities. This, combined with the increasing competition for providers’ dollars and funding challenges, will hasten consolidation in the market.
There have been some early signs that this consolidation is starting to bite, including MaxQ.ai’s pivot away from the medical imaging AI market, Nanox’s acquisition of Zebra Medical Vision and more recently Tempus’ acquisition of Arterys. Such headlines are likely to become more common in 2023 and beyond as it becomes increasingly difficult to compete with the established cohort of market leaders and their more sophisticated solutions. This impetus is also likely to give rise to other trends; vendors turning away from radiology to other markets where their capabilities might be in higher demand, or they are not hindered by the same regulatory hurdles. Vendors may, for example, look towards pharmaceuticals and drug discovery.
Ultimately, these shifts in medical imaging AI could leave the market drastically changed in several years. Fewer vendors, with broader capabilities, smaller vendors acquired and subsumed by larger market leaders, healthy reimbursement, and true mainstream adoption, even several unicorns traded publicly. Regardless of these potential changes, the foundation of the successful companies will remain the same. The companies that have success in the future will still be the ones that can offer, evidence, and deliver clinical value. Essentially, vendors capable of delivering on AI’s fundamental promise will continue to thrive.
About Signify Premium Insights
This Insight is part of your subscription to Signify Premium Insights – Medical Imaging. The content is only available to companies that have subscribed to this paid-for service. To view other recent Premium Insights that are part of the service please click here