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 Dr Sanjay Parekh
Last month, Bayer, a company best known in radiology for its contrast agents and radiation dose management solutions, announced that it was launching a new AI platform.
Dubbed Calantic, the new offering aims to make it easier for providers to deploy and integrate AI solutions, giving providers easy access to tools that aid radiologists in prioritising cases, detecting lesions, quantification, and productivity.
The solution is a cloud-hosted, vendor neutral platform and bundles together tools to offer suites dealing with specific workflow tasks, based around particular body regions or procedures. Initially, these suites will focus on thoracic and neurological disease.
AI platforms are a growing segment of the medical imaging market, but it is also one that is well catered for. The key question is therefore: Will Bayer’s platform offer enough to stand out in such a crowded market?
The Signify View
Although use of medical imaging AI is growing, there are still a number of barriers stymieing the adoption of the technology. Among these is the challenge that providers face in deploying AI tools and integrating them into their hospital IT systems. This barrier is increasingly being eroded, however, with a number of AI platforms in particular facilitating the adoption of these tools.
This is the market into which Bayer has launched Calantic. On the face of it, gaining traction in such a market could prove difficult. There are, after all, a large number of players all vying for success, with platform offerings now available from a host of different companies, from large imaging IT vendors such as GE and Siemens looking to make AI provision a central element within their PACS, to burgeoning independent AI vendors looking to utilise platforms to support their own algorithms such as Aidoc, to third party marketplace providers, such as Blackford Analysis, Terarecon and Nuance. Add to this a new wave of specialist platform providers from the likes of DeepC and Carpl.ai, and it’s clear the segment, though young, is already very crowded.
Another hurdle facing Bayer is that its Calantic platform, offers a lot of the same capability as its competitors, which might make it difficult for potential customers to choose the new offering over other more established players. It, like those current solutions, aims to solve the last mile challenges of deployment integration and orchestration, with similar focuses, such as supporting triage and case prioritisation as well as detection and quantification. Even the vendor’s workflow suites may struggle to stand out, offering only a handful of partner solutions to its users, compared to the high double digits, or even hundreds offered by some of its competitors. Its neurology package for example is focused on ICH and LVO, using algorithms from Avicenna. This same capability is, however, also available on competing platforms. Arterys’ Neuro AI platform, for example, offers Avicenna’s stroke solutions, while also offering additional capability, such as brain perfusion analysis, brain volumetric analysis and lesion quantification tools in one package.
A Terminal Diagnosis?
Such challenges may, at first, seem terminal, but despite the competition and the present limitations in content, there are some features that could make Calantic an attractive option. The vendor, for instance, allows customers to test the hosted algorithms before they must commit to purchasing them. Although a relatively straightforward capability, it is very valuable for providers which are trying to determine which are the best solutions for their own, unique situations, and will help them ensure they select the most appropriate tool for their needs. This capability will only become more valuable to providers as the number of tools available on Bayer’s platform increases, allowing providers to better choose between algorithms offering comparable capability. It also, however, is not unique to Bayer’s platform.
There are also other, end-user focused features which could give Calantic a potential edge over some competitors. The value of offering many solutions to a radiologist will be mitigated if the radiologist isn’t able to use them easily. Calantic offers the ability to see, and interact with, the results of AI within a provider’s PACS, streamlining a radiologist’s workflow. This can be a significant advantage versus some AI vendors, ensuring a radiologist’s attention isn’t diverted as they open another programme, or even log in via a browser to access a tool’s result. This integration also extends to the report. Unlike many other solutions, Calantic offers what it calls Auto Documentation, which, Bayer says, populates the report with AI descriptive findings, saving the radiologist time, who would otherwise have to decide what information to include and copy it or type it themselves.
This is another feature that will also grow in utility as AI use becomes more sophisticated and widely adopted. The number of quantitative findings AI tools will be able to draw from medical images is going to increase, making it increasingly important for this information to auto populate reports, and not rely on radiologists transferring large numbers of quantitative findings from programme to programme. However, Bayer is not alone in trying to facilitate the “last-mile” challenge of integration and orchestration, with a host of competitors also at various stages of competency in this regard.
Relinquish the Reins
Automation, however, can be a mixed blessing. While it can streamline tasks and save time, the automation of processes can also allow errors to be entered unchecked, leading to difficulties down the line. Automating radiological processes can also rob radiologists of their sense of control. Bayer attempts to sidestep this issue in its Calantic platform with an accept or reject feature, giving radiologists the ultimate control over what is entered into the record, as well as ensuring they are engaged, and not overly reliant on automated tools. Again, not unique, but a necessary workflow step that should aid adoption and win over users.
Aside from well-thought-out features, Bayer’s pedigree in radiation dose management tools also confers some advantages that many of its peers lack. The use of Bayer’s dose management tools in screening programmes, potentially beats a path for Calantic’s use alongside them; a move that would allow the detection of incidental findings and bestow the platform with considerable population health credibility. Bayer’s experience in utilising analytics may also drive interest. Such capability, if applied to Bayer’s platform could be particularly advantageous, enabling providers to better assess the impacts of the solutions they have deployed. This sort of evaluation could be valuable, not only guiding a provider in the selection of an alternate if its initial choice fails to meet its needs, but also informing how best to utilise a deployed algorithm. In such a way, Calantic, or another solution that offered analytic capabilities, could ensure that hospitals both selected the right algorithms, and maximised the value they derived from it. However, traction for advanced dose management platforms has been an uphill battle for the firm (and others), given its core sales channel via contrast agent and injector supply. While this is changing and the US has been a success story for Bayer more recently for radiation dose management, its presence in the market compared to leading imaging IT and imaging modality vendors remains modest.
Whether these features are enough to differentiate Bayer’s Calantic from the rest of the players on the market remains to be seen. Bayer has created a platform that brings together many attractive features from across the market into a single sophisticated package. Such a package will also be bolstered by Bayer’s experience and existent install base, which, assuming the vendor is able to convince its customers of the merits of the adoption of AI, could allow it to quickly establish itself as one of the leaders in the burgeoning platform sector. It would still have to negotiate potentially difficult relationships with imaging IT vendors, some of which may look to prevent third-party platforms, but this difficulty is likely to be overcome if there is enough appetite among customers.
A more immediate concern is the limited number of solutions hosted on the platform. While it is important that solutions are well curated and well-integrated, the paucity of options for providers renders it, at present at least, a tough sell. Fortunately for Bayer, this is a problem that can be solved. If Bayer can do so quickly and effectively, giving providers more options and fleshing out its neurology package and thoracic packages to increase their utility then it can be a compelling choice for a welcoming customer base. Moreover, with an ongoing focus on contract consolidation and a growing digital dose management presence in the US, some customers may look to it as a safe pair of hands versus a less known market entrant. If not, it risks becoming an innovative also ran in a crowded and competitive market.
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