Tag Archives: Breast

Signify Premium Insight: iCAD’s Woes Require More Being Done with Less

iCAD have long been among the frontrunners in the breast imaging AI market. Although breast imaging is one of the longest established use cases for medical imaging AI, the vendor has, in the past, still performed strongly compared to its peers.

This strength continued during 2020 and 2021 when the Covid-19 pandemic was most prevalent, which, in many cases, caused breast screening programmes to be paused. Despite such a setback the vendor was able to maintain, if not increase revenues.

However, such a run has ended, with the vendor’s latest results showing a decline of more than 20% in its detection product revenue. Of more concern, this, combined with results from its other segments, means that overall, in 2022, iCAD’s expenses were 21% higher than its revenues.

Signify Premium Insight: US-FDA’s Standardised Appeal

Last month the US-FDA added a new rule to the Mammography Quality Standards Act, adding detail on the reporting of breast density in screening programmes.

The new guidance means that, from September 2024, breast screening centres will be required to notify patients about the density of their breasts, a physiological feature which can influence the accuracy of mammography and a risk factor for developing breast cancer. Providers will also have to convey this density information in specific, non-technical language.

The new rules also standardise the reporting of breast density, establishing four categories to identify breast density levels in the mammography report. Such federal guidance has been on the cards for several years, so what can be expected now it is enacted?

Signify Premium Insight: AI’s Personal Approach to Breast Imaging

Screening programmes are among the medical imaging use cases where, at least for the majority, there is little urgency. As such, they were one of the first casualties of the cancellations in response to the developing Covid-19 pandemic in 2020, as providers sought to stem the spread of the virus. While the rationale for such action was sound, now, almost three years later the consequences are being keenly felt.

According to Bhvita Jani, Medical Imaging Research Manager and co-author of Signify Research’s Breast Imaging Report – 2022, these consequences are having a significant impact on the breast imaging market, although there are some technologies that are coming to the aid of providers looking to address the challenges.

Signify Premium Insight: RadNet is Full Screen Ahead with AI Acquisitions

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.

Last week, imaging services provider RadNet announced it had acquired Dutch artificial intelligence developers Aidence and Quantib. Aidence is a radiology AI vendor which specialises in the development of clinical solutions for pulmonary nodule detection and lung cancer screening, while Quantib is an AI developer which, offers solutions for prostate cancer and neurodegeneration.

Aidence and Quantib will build on the AI capability RadNet acquired when it purchased Nulogix and DeepHealth in 2019 and 2020 respectively and leaves the American company with considerable potential in the field of cancer screening, an area in which RadNet will increasingly focus.

The Signify View

Many investors are impatient creatures. This impatience often serves them well, allowing them to spot a suitable opportunity and make a tidy return in a relatively short time frame. For the companies they invest in, however, this impatience can force them to alter and adapt their strategies, and ensure they must react to opportunities.

This flexibility is particularly important if a company is running at a loss of around $4-$5m dollars a year, against a figure of just $13m raised from venture capital investors. This is the situation that Aidence and Quantib found themselves in. Both vendors had developed promising AI technology but, with AI adoption in its infancy, commercialising and capitalising on this technology is still something of a challenge.

RadNet believes that in acquiring Aidence and Quantib, it will be able to address this issue. In combining Aidence and Quantib, with the AI capability it already possesses (via DeepHealth) as well as the extensive network of more than 290 outpatient radiology centres, the American provider believes it can create a powerful cancer screening network. It will also look to bolster the capability it has now assembled in breast, lung, and prostate, with colon cancer screening, to be able to offer a screening solution that detects around 70% of cancers that are imaging detectable at an early stage. Currently breast is the only cancer that is widely screened for, but RadNet’s strategy will see it increasingly push for greater levels of screening of these (and other) cancers, aiming to see them become as mainstream as breast screening. As this happens, RadNet will be able to leverage both its new and older acquisitions to be able to offer and deliver multi-cancer screening, capitalising on the comparatively vast numbers of patients that this would entail.

As a strategy this is shrewd, as offering screening is an additive rather than binary development. RadNet can, in the immediate term, use its breast AI tools for breast screening, offering lung and prostate scans as diagnostic clinical decision support tools, but, deriving increasingly greater proportions of revenue from them as these screening types become more common.

Screening’s Scale

The fact that potentially vast numbers of patients will be imaged under screening programmes also reinforces RadNet’s decision to turn to AI tools. As a provider, RadNet will be competing with other outpatient imaging centres, as well as acute sites. This growing competition means that investments which give a provider such as RadNet even a slight competitive advantage will be worth considering. Bringing separate companies and their technology to create an efficient screening solution at present, when screening programmes outside of breast cancer are nascent, is somewhat speculative, but it will give RadNet an advantage, enabling it to drive forward and help create the market for such practices.

Even without the additional impetus given by screening programmes, prioritising the acquisition of AI capability is a sound strategy for RadNet. Both of RadNet’s latest acquisitions, but particularly Quantib’s solutions, are focused on time consuming medical imaging examinations. These involved examinations are comparatively expensive requiring large amounts of radiologist time. Radiologists are among RadNet’s most significant expense, with the provider, according to CEO Howard Berger, spending around 20% of its global net revenue on this resource.

AI solutions which streamline the reading workflow and enable radiologists to be more accurate and efficient, can represent a significant saving for the outpatient imaging provider, and therefore give it increased pricing flexibility and bolster its competitive credentials. This will both allow it to compete directly with other providers, as well as establish itself as an outsourcing option for larger providers for these highly time-consuming exams. If acute hospitals realise that an AI-equipped RadNet can perform and read imaging exams, take a share of the revenue, and still undercut acute centres, then those centres are likely to simply outsource some imaging procedures to them, rather than trying to compete with them on cost. This outsourcing looks particularly attractive for high volume and time-consuming examinations, such as brain MRI and prostate MRI exams.

Sold on Strategy?

One of the more unexpected aspects of RadNet’s acquisition of Aidence and Quantib is the provider stating its ambition to expand its presence both in North America and Western Europe, and positioning itself as  an imaging IT and AI vendor as well as provider. RadNet intends to use its newly curated multi-cancer screening AI tools as a potential lead for its self-developed eRad products. This will prove a challenging task, with the imaging IT market increasingly saturated  in mature regions making organic growth hard to come by. Most providers are already committed to lengthy deals with better established vendors, meaning that RadNet or other vendors eying a provider’s custom will, in most cases, have to wait several years before even having the opportunity to tender for their business. What’s more, providers are becoming less likely to “rip and replace” their existing imaging IT platform. Challenger vendors therefore need to present a significant value proposition that makes the time and cost a worthwhile investment for a provider.

RadNet, for its part, has pushed acquired centres to use its eRad products, regardless of the incumbent solutions or deal lengths. Although this has worked to date, it is less likely to translate to outside of outpatient imaging in international markets.

RadNet does have some advantages, though. As a provider itself, the AI solutions and eRad products will be used internally, and meet the company’s own need, as such, revenue brought in by the sale of the imaging IT software will, in the near term at least, be supplementary to RadNet’s core business. It is deriving another revenue stream, even if small, from already sunk cost.

If RadNet can bring its DeepHealth, Aidence and Quantib acquisitions successfully into a multi-cancer screening solution it could also hold a seriously advantageous position in the cancer screening market. While there are some AI developers that have focused on breast screening solutions, such as Volpara Health, RadNet will be among the first to offer a more broadly capable package targeting multiple cancers. This could put it in a particularly strong position if screening for other cancers becomes more common. Guidance around lung screening in particular is becoming more encouraging, with bodies broadening the criteria for which screening can be used and new studies (the latest taking place in France, funded by the Ministry of Health and the National Cancer Institute), all helping to build momentum for screening. RadNet’s position makes it look likely to capitalise on any rapid increase in screening, from both a vendor and a provider point of view. In fact, RadNet’s unique position may even enable to become a driving force for other cancer screening programmes beyond breast cancer to be adopted. However, whether it commits to such a plan highly depends on whether it has the appetite to champion such initiatives at a legislative level whilst growing its AI division.

A Matter of Timing

Although RadNet’s move to acquire Aidence and Quantib harbours some risk in that it is counting on cancer screening programmes growing in prevalence and becoming more widespread, it is a move that makes sense. The AI market is maturing and, as highlighted by MaxQ’s failure last month, developers that have been surviving off shrinking venture capital (especially in Europe), and which have failed to gain significant market traction will find themselves in increasingly precarious positions. RadNet has therefore chose an appropriate time to make such acquisitions. This timing will look particularly well-judged should the cancer screening market beyond breast cancer, take off as the provider expects.

In their decisions to sell, Aidence and Quantib also acted prudently. Their VC investors have enabled them to grow into companies harbouring competent, regulator approved solutions, but, on their own, both would face uphill battles trying to commercialise their solutions, a process that could be difficult to sustain given the losses they are making. To sell now allows the developers to, as part of a larger company, build on these foundations and become part of a more valuable whole.

Many AI developers that are finding further VC funding hard to come by, and watching their bank balances dwindle, will be hoping to be as fortunate. The AI market continues to move on, as it does, more companies will follow RadNet’s lead and decide it could well be the right time to jump on board.

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

Signify Premium Insight: Breast Imaging: A Market Set for Change

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.

The nature of the breast imaging market makes it unlike almost all others. The fact that it is a market centred around national guidelines and screening programmes, which were almost universally paused as part of providers’ responses to the coronavirus pandemic, means that it has been disrupted more so than most others.

This disruption, and the long-term effect it is set to have on the market, as well as the impact of developing trends are detailed in the recently released Breast Imaging – World Market Report 2021.

For providers, the response to this disruption is likely to be varied.

“One statistic in the report,” says Senior Research Analyst and co-author Graham Cooke, “is that the UK’s National Health Service (NHS) expects it to take ten years to clear the backlog.”

“Because of this, if a healthcare provider does have the money to invest in new equipment and upgrade to new technologies then dealing with this backlog provides an incentive to do so. However, it could be the opposite. Breast imaging clinics could well have been stretched with regards to available budget over the previous year, and with restricted funds, they could continue to use the ageing installed base of equipment until there are funds available to replace them.”

 A Digital Direction

For some of these providers, however, the answer may not necessarily lie in the replacement of breast imaging hardware, but could focus on maximising the use of existing systems.  “That is really where AI could come in,” continues Cooke. “AI isn’t just about finding abnormalities on a mammogram that the human eye may miss, more practically some AI tools could help radiologists detect abnormalities more quickly, or rapidly give some cases a green light where no further attention is needed. This will mean that radiologists can focus their time on more urgent cases.”

While these tools could relate directly to image analysis, in practice, many of the solutions that are set to be of most use in clearing the backlog of postponed breast imaging exams focus their efforts elsewhere. Tools that facilitate operational workflow improvements or triage tools that help prioritise suspicious scans in the worklist are among those that offer the greatest benefit.

“Another interesting tool that will help in this regard is personalised screening and having AI assess a woman’s individual cancer risk, including factors like family history of breast cancer, genetics, lifestyle factors which make some people more at risk, or physiological factors like dense breast tissue.”

This will help providers manage their screening programmes, prioritising those who are most at risk, while individuals at lower risk of breast cancer can wait longer between mammography appointments, thereby reducing the provider’s workload.

This, however, will not happen immediately, cautions Senior Research Analyst Bhvita Jani. “At the moment there are still three main barriers to the widespread use of AI in breast imaging. Firstly, until it is included and recommended in national screening guidelines it is not going to make as much impact as it should.

“Secondly there is a lack of representation in the validation studies. There is, for example, an underrepresentation of some ethnic minorities which makes it impossible to extrapolate the findings to women of different backgrounds.

“The third problem is reimbursement. Unless AI is reimbursed the only way that providers would be convinced to use it is if it will reduce their costs.”

The breast imaging AI market is set to grow rapidly as the barriers to its adoption are overcome

Starting Screening

Another factor that could dramatically impact the breast imaging market is the establishment of screening programmes in countries that don’t yet have them. The growth potential for different vendors in the breast imaging market is highly dependent on the country dynamics.

“China, for instance, is quite a price-sensitive market,” notes Cooke, “so even if it were to introduce a screening programme, that screened millions of women, providers would overwhelmingly purchase their systems locally. Not only do Chinese vendors have lower price points than international vendors, but Chinese hospitals often have partnerships with local vendors. Going with a local brand will also often be seen favourably, compared to using an American or European brand for example.”

Jani adds: “But, this would only be the case for 2D mammography systems. Any demand for 3D systems will still primarily be captured by global vendors as Chinese vendors don’t extensively offer 3D systems.

“In some areas that are on the cusp of establishing screening programmes global vendors will be preferred. In the Middle East region, for example, where brand reputation is very important. Screening programmes in other areas could also benefit global vendors, including several CIS (Commonwealth of Independent States) countries. However, within this market, there is a stronger preference for low to mid-range breast imaging systems.

“It is worth noting though, that when it comes to large government tenders, price is often the most important consideration. So, companies that are more flexible when it comes to price negotiations have tended to do well. One company that has been particularly successful is Fujifilm. It has increased its share of the mammography market, through, in part, its aggressive pricing strategy.”

 Demand for More Dimensions

For those countries that already have established breast screening programmes, there are still some large and impactful changes on the horizon.

“One of the biggest unanswered questions in the breast imaging market is when will 3D mammography be used outside of the US for screening, which is the only country to have so far really embraced the technology,” Jani explains. “Our prediction is maybe three to five years for Western Europe, which looks set to be one of the next markets to adopt 3D for breast screening.

“This rollout of 3D mammography can only take place when there is enough provision to read the 3D scans, with limited radiologist capacity being a challenge. Because of this, the rollout of 3D will likely go hand in hand with the increased adoption of AI, which will help providers make the transition by reducing the time required to read 3D scans.”

Digital Breast Tomosynthesis is forecast to grow strongly over the coming years

There could also be other technology changes afoot. “There are a number of different emerging breast imaging technologies becoming available,” notes Cooke, “but mammography is still the dominant, unthreatened modality by far.”

“Other modalities, such as ABUS or Cone Beam Computed Tomography have advantages compared to mammography in some ways, but it is difficult for them to penetrate the breast imaging market. “These technologies are not designed to replace mammography but are instead meant to supplement it. They must work alongside mammography to be part of patient pathways. If you consider automated breast ultrasound, for example, that has taken some 20 years or so to really be accepted, and even now there are only a handful of vendors that offer it.

“So, we could see some of these smaller and more niche breast imaging vendors getting acquired by larger breast imaging vendors. There are already examples of this, with Hologic, which dominates the mammography market, diversifying their hardware portfolio to include breast ultrasound, by recently acquired Supersonic Imagine, whose technology should complement Hologic’s own mammography systems.”

Women’s Priorities

There are other changes that will affect providers’ purchasing decisions in the nearer term. For screening mammography at least, screening centres are in competition with one another, in particular in the United States, for footfall of eligible women. Because of this, vendors will happily pay to purchase new and higher-end systems, with features which focus on the patient experience and comfort, if they will attract more women into their centres.

“One-way providers are encouraging more women into their screening programmes is by focusing on enhancing comfort levels, and reducing discomfort and pain,” notes Cooke. “That could have a large role in influencing what systems providers actually purchase.

“This is a prominent trend and as a result, instead of only marketing their systems to providers, breast imaging vendors are targeting women directly. The preferences of these women could then influence what systems providers purchase.”

The Market Matters

These sorts of strategies are only viable in countries where screening programmes are already established. While there will be effort expended to encourage women to participate in screening programmes, for vendors and providers in these countries the priority will be on improving the patient experience and the accuracy of first-time right screening and reducing the rate of rescans and recalls.

“Over the coming years, there will be more focus on personalised screening and risk-based screening, as well increased variation in the breast imaging modality used and the frequency of scans,” details Jani. “These changes could also be very much influenced by AI’s development in this market.

“In developing countries, the creation and establishment of screening programmes will be the biggest change. But this is itself dependent on the subsequent treatment plan that is available afterwards is a diagnosis is made”.

“Overall, though, the next 10 years could really shake up the breast imaging market. New emerging breast imaging technologies could gradually have increased market adoption where there are established screening programmes. Meanwhile there will be some opportunity for niche vendors to increase the uptake of their products and get their systems installed as part of new screening programmes where guidelines are yet to be established.

“It is, after all, easier to penetrate a new market with a lack of guidelines and recommendations, than it is to try and break a well-established market like the US right away.”

The breast imaging market is set for change. Technological developments, primarily in breast AI, will play a part in enabling this change and drive increased uptake of more advanced breast imaging hardware solutions. With current challenges such as the significant backlog of women requiring screening, a fundamental need to reduce the number of false positives or negatives and the need to increase early detection and more accurate diagnosis all being factors that will help drive this movement forward.

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