AI in healthcare: Are you tracking the right competitors?

When Clarivate developed the AI50, we wanted to look beyond the organizations most visibly associated with artificial intelligence (AI). Using patent data, we examined which companies were building significant bodies of AI invention, both through the scale of their contribution and the importance of AI within their wider inventive activity.

What was surprising was the way these companies applied their inventions. Technologies developed by computing, semiconductor and digital-platform companies were appearing in highly specialized fields, including healthcare.

That prompted us to examine the life sciences-related inventions within the wider Clarivate AI50 analysis. One finding immediately stood out: only two pharmaceutical companies held more than two AI inventions ranked among the strongest 0.5% globally, while more than 50 medtech and biotechnology organizations met the same threshold.

The finding reveals where organizations are securing AI-related rights and why those positions may matter to future healthcare competition. Interpreted as patent intelligence, the data can help IP teams identify the companies and capabilities likely to influence future products, partnerships and freedom to operate (FTO). Placed alongside scientific, commercial and market evidence, it creates a more complete view of where strategically important AI capabilities are developing.

The data bring an unexpectedly diverse competitive field into focus. Alongside established medtech companies are focused life sciences specialists and businesses from industries that include computing, semiconductors, consumer technology and digital platforms. Some hold AI inventions with direct applications in medical imaging and other clinical technologies.

For IP teams, this exposes a weakness in competitor monitoring built around industry categories alone. A company developing an enabling technology for a future healthcare product may not sell a comparable product today or operate primarily within the life sciences sector. As AI becomes more deeply integrated into medical devices, diagnostics and clinical workflows, IP intelligence needs to follow the technology across industry boundaries.

Why medical technology is so visible in the patent data

Medtech companies, including GE HealthCare, Medtronic, Siemens Healthineers and Philips, are among the most prolific owners of the highly rated AI inventions identified in our life sciences analysis.

It would be tempting to turn that result into a simple comparison between medtech and pharmaceutical companies. But that would miss the more interesting point: the scale of protected AI technology being assembled around healthcare systems.

Medtech companies are building substantial patent portfolios across the technologies that connect AI with medical devices, diagnostics, imaging and clinical workflows. For IP teams considering where healthcare AI is heading, those portfolios provide an important signal: they show where relevant capabilities are developing, which organizations are securing rights around them and where those rights could influence future products.

Pharmaceutical AI activity may appear differently in patent data or be protected through trade secrets and other means. Viewed alongside broader scientific, commercial and market evidence, the depth of medtech patent portfolios provides IP teams with a clearer view of the organizations and technologies shaping the future healthcare landscape.

AI is becoming part of the clinical system

Elekta offers a useful example.

The company holds four of the highly rated AI inventions identified among the medtech specialists in our analysis. Its work brings AI together with imaging, real-time data, treatment planning and the equipment used to deliver radiotherapy.

As Vandita Chandrani, Head of IP at Elekta, explains:

“Our commitment to precision radiotherapy pushed us to innovate early with AI. By harnessing real-time data, advanced imaging and adaptive control systems, we’re optimizing accuracy for more responsive and personalized treatment. Integrating AI across our linacs, imaging and treatment planning helps us continue to push the boundaries of what’s possible in radiation therapy.”

The significance lies in how AI becomes part of an interconnected clinical system. Relevant IP may extend across the algorithm, the data it interprets, the imaging technology, the control mechanism and the equipment acting on its output. Different organizations may own rights across that technical architecture, including suppliers, collaborators and companies operating outside the immediate market.

IP teams need to understand the products they compete with, the enabling technologies those products depend on and the organizations building patent positions around those technologies.

Focused specialists can matter before they become established competitors

The analysis also brought a varied group of focused specialists into view, including Anumana, Lunit, Vignet, Venus Medtech, Sysmex, Promation, OD Vision, Lepu, LedgerDomain, Insitro and Athena Eyes.

These organizations differ considerably in their technologies, markets and stages of development. Their presence in the data does not mean they will all become major players. But while revenue and market share show who has already achieved commercial traction, patent data can bring an organization into view while its technology is still developing.

For IP leaders, that suggests a different question from the familiar: Who competes with us today?

The more forward-looking question is: Which organizations are developing technologies that could influence what we build next?

That shift matters. By the time a specialist becomes an established competitor, the opportunity to partner, license, acquire or develop an alternative may have narrowed.

Relevant capabilities are emerging across distinct industries

Industry limitations become even less useful when healthcare-relevant inventions originate outside the life sciences sector.

Our analysis also found highly rated AI inventions with life sciences applications associated with Tencent, Alphabet, NVIDIA, GE Aerospace, Baidu, Samsung, Intel and Reebok.

Across this diverse group, the relevant inventions draw on capabilities including computing, machine learning and data processing that can be applied to healthcare. Their strategic relevance will vary by product and technical area, but their patent positions may influence the systems organizations develop across the life sciences sector.

One NVIDIA invention^[1], for example, concerns techniques for training neural networks to detect and segment objects in medical image data. It illustrates how healthcare-relevant IP can originate with a company more commonly associated with computing infrastructure than clinical technology.

Depending on the context, a company from an adjacent industry may be a supplier, development partner or relevant rights holder. Its enabling technology can shape a healthcare product even when another organization brings the complete product to market. Together, these relationships create a more interconnected competitive environment in which technical relevance extends across established industry boundaries.

Follow the technologies that matter

A broader view only helps when it is tied to the organization’s strategy. IP teams can begin with the capabilities future products are likely to require, using those priorities to determine which semiconductor, software, computing infrastructure and digital platform companies warrant closer attention.

This creates a new starting point for competitive intelligence: begin with the technologies that future products may require, then trace those technologies to the organizations that develop and protect them.

Four decisions that a wider view can improve

Following technology across industries can strengthen four connected areas of IP strategy.

Competitive monitoring

Monitor companies that sell comparable products and organizations that develop capabilities relevant to future products or clinical workflows. The aim is to distinguish meaningful technical adjacency from background noise and identify which technologies, partnerships or patent positions could influence the organization’s plans.

Build, partner, license or acquire

Focused patent portfolios can indicate that organizations are developing concentrated expertise around a technical or clinical problem. Patent strength is one input alongside scientific validity, technical performance, regulatory progress, commercial position and organizational fit. Used carefully, it can help IP and R&D teams decide where deeper investigation is warranted and ensure less obvious specialists enter the build, partner, license or acquire discussion.

FTO

A digitally enabled medical device may combine equipment, sensors, data processing, trained AI models, software interfaces and control systems. The rights associated with those technologies may be held by organizations that sit outside the established healthcare competitor group. Mapping the technologies that enable a product can help IP teams identify relevant rights holders and areas that may require closer investigation. Legal analysis can then be based on a fuller understanding of the technologies, organizations and rights that may be relevant to the product.

Portfolio development

A cross-industry view can show where invention is concentrating and where room for differentiation may remain. It can prompt sharper questions about alternative technical approaches, combinations of capabilities and the transfer of existing expertise into emerging areas. Scientific, commercial and competitive investigation is still required, but patent intelligence helps direct that work toward the most consequential questions.

A different way to draw the competitive map

The central finding is that healthcare AI connects organizations that conventional industry categories keep apart. Established medtech companies, focused life sciences specialists, and businesses from adjacent industries may operate in different markets, yet their inventions can form part of the same technology domain.

That changes how the competitive map should be drawn. Organizations may enter the ecosystem as suppliers, collaborators or owners of enabling IP long before they become direct competitors. A competitor list based only on current products risks overlooking the capabilities that future healthcare products will depend on.

Apply the Clarivate AI50 lens to your technology domain

The life sciences analysis offers a sector-wide view. Applying the same approach to a specific portfolio, technology or product roadmap can help identify relevant patent positions, emerging concentrations of invention, less obvious rights holders and organizations that may warrant closer investigation.

Explore the full Clarivate AI50 to see which organizations are building significant positions in patented AI invention.

Read the Clarivate AI50 report

See how the same approach could be applied to your portfolio or technology domain.

Talk to our team

About the analysis

The Clarivate AI50 examines two dimensions of organizational AI invention: an organization’s share of high-strength AI inventions and the importance of AI within its wider inventive activity. The life sciences findings in this article are drawn from a closer examination of relevant inventions within that wider analysis.

Invention strength was assessed using the Derwent Strength Index. The comparisons in this article refer to AI inventions ranked within the strongest 0.5% globally.

Patent data capture inventions that organizations choose to protect. It does not capture all AI development or use, including activity retained as a trade secret, released through open-source channels or not protected through patents. The findings should therefore be considered alongside scientific, technical, legal, commercial and market intelligence.

[1] US12462377B1, DWPI title: “Processor for use in medical imaging system for training neural networks for detecting objects in image data for purposes of analysis using AI, has neural networks that are trained to generate second segmentation based on use of first segmentation as supervisory signal.”

AI Patents Pharmaceutical