It began not with a bang or a singular, disruptive event, but with something far more subtle—more undercurrent, really, than upheaval.
The usual suspects were in place: lingering trade wars, geopolitical volatility, deglobalization, economic divergence, multipolarity, and, of course, nature’s fury. Amid all the turmoil, though, was the whisper of change—a shift from the obvious digitization of years past to ambient artificial intelligence (AI) integration, hyper-personalized treatment, and interconnected health ecosystems.
With the pandemic’s ripples finally stilled, medtech firms made fundamental changes to their business operations, commercial structures, care delivery models, and portfolios this year to better align with current trends and rapidly expanding markets.
“…we’ve seen a lot of investment go into markets like robotics that are growing well above 5%,” EY Global Medtech Leader John Babitt noted during a mid-October media briefing. “Diabetes is another area, structural heart, certainly pulsed field ablation, and even some technologies going into neurovascular. Those [are] a handful of the areas…”
AI, however, was the clear winner this year. Medical device companies doubled down on the technology as it became an increasingly integral part of their solutions. AI’s role in diagnostic imaging continued to grow, with GE HealthCare maintaining its perch atop the FDA’s list of AI-enabled medical devices (90% of which are radiology and image-focused), and embedded AI began shifting cardiovascular solutions from monitoring physiology to detecting changes early enough to guide treatment.
“Increasingly, AI is enabling devices not only to capture and interpret information but to anticipate changes in a patient’s condition, personalize treatment and, in some cases, act with decreasing levels of human intervention,” EY’s Pulse of the MedTech Industry Report 2026 states. “For medtech companies, this means the focus of value is migrating from the device to the intelligence around it, with real consequences for how they compete, earn revenue, and organize.”
The new (artificial) reality is taking shape.
AI Everything, Everywhere All at Once
It was a natural evolution, of sorts.
Well, more like an unnatural evolution, really.
Its genesis began more than six decades ago in a Massachusetts Institute of Technology laboratory, where a bunch of computer source code morphed into a Rogerian psychotherapist who could interact with humans. That codebase consultant not only processed code, it also conversed with people, engaged them in dialogue, and even had a name.
Taking its moniker from George Bernard Shaw’s heroine in “Pygmalion,” ELIZA was the world’s first official chatbot (though it was not identified as such at the time; rather, it was deemed a “natural language conversation” program).
“Seen in the coldest light possible, ELIZA is a translating processor in [Saul] Gorn’s sense; however, it is one which has been especially constructed to work well with natural language text,” ELIZA creator, computer scientist/MIT professor Joseph Weizenbaum, wrote in a January 1966 paper published in Communications of the ACM. “It is useful to remember that the ELIZA program itself is merely a translating processor in the technical sense. A large part of whatever elegance may be credited to ELIZA lies in the fact that ELIZA maintains the illusion of understanding with so little machinery. But there are bounds on the extendibility of ELIZA’s ‘understanding’ power…”
Those bounds gradually dissipated with each successive generation of translating processors: PARRY (1972), for example, modeled the behavior of a paranoid schizophrenic (“ELIZA with attitude” as described by its creator); Racter (1984) and ALICE (1995) went further, generating increasingly sophisticated text-based conversations; while Cleverbot (2008) promised something even more ambitious—the capability to learn from its conversations. Virtual assistants Siri (2011) and Alexa (2014) brought daily multi-functionality (setting reminders, ordering groceries, controlling smart devices) into the mainstream.
Healthcare’s artificial intelligence (AI) evolution followed a similar trajectory, starting with MYCIN (early 1970s), a Stanford University program that recommended treatments for certain infectious diseases. AI pioneer Alan Newell called the tool the “granddaddy” of all expert systems, and “the one that launched the field.”
Ascending from that launch was software such as COGNIsoft (1997) and Beating the Blues (2003), which brought structured cognitive behavioral therapy to end users. Natural language processing took matters a step further, enabling platforms such as Ada Health and Babylon Health to guide patients through preliminary symptom checks. The COVID-19 pandemic accelerated AI’s natural progression: Governments and clinicians used the technology to track and contain the virus, while machine-learning algorithms and deep learning systems helped speed vaccine development and delivery.
Since the pandemic, healthcare-focused AI has evolved at warp speed, infiltrating nearly every corner of medicine—from drug development and device design to manufacturing, testing, and packaging. AI-enabled digital twins, for instance, can help engineers design safer, more inclusive devices by simulating the way products behave under real-world conditions across diverse patient populations. Similarly, AI-powered predictive maintenance can keep production lines running smoothly by spotting problems before they become costly disruptions. The technology can also unravel the root causes of test failures far faster than traditional statistical methods, detecting complex connections across materials, software, and mechanical systems that might otherwise go unnoticed.
“Artificial intelligence is everywhere in healthcare and life sciences. AI has helped medical device manufacturers maintain regulatory readiness as devices are updated to new designs and best practices. It has significantly reduced the time required for research and development of new drugs, treatments, and devices. And AI has even been used to create digital twins of the human body to map tumors and help treat cancer,” a NVIDIA Survey Report concluded in February. “In short, AI has been truly transformational for the industry. AI has been growing into healthcare and life sciences for several years, and the maturity curve has hit an inflection point in the last year.”
Inflection point indeed: Medtech’s “Everything, Everywhere, All at Once” moment has arrived—and AI is at its core.
The technology reached a turning point this year, as medtech firms shifted away from point solutions and standalone tools toward deeper integration across clinical and device ecosystems.
Case in point: In May, Siemens Healthineers secured U.S. Food and Drug Administration (FDA) clearance for six new systems in its Artis interventional imaging portfolio, featuring the Optiq AI imaging chain, which leverages artificial intelligence to analyze and optimize image data in real time. Optiq AI uses deep learning to reduce noise across 2D imaging modes—including fluoroscopy, acquisition, and digital subtraction angiography—delivering consistent, high-quality visualization during interventional procedures.
The company’s Varian business earned CE mark authorization the same month for its AI contouring capability in the Eclipse treatment planning system. The integrated software automatically contours more than 200 predefined structures from CT and MR images, including organs of interest, lymph nodes, and known brain metastases. It also provides a standardized starting point for clinicians to review and refine contours, thereby helping improve consistency and reduce manual effort in radiation therapy planning workflows.
“Building on our experience with AI contouring, we are integrating this capability into radiotherapy workflows, enriching treatment planning with Eclipse, and helping to broaden actionable access to AI-enabled precision consistently throughout the entire workflow,” Rebecca Schuster, head of Cancer Therapy Imaging at Siemens Healthineers, said upon CE mark receipt.
Similarly, Medtronic expanded its AI-powered clinical ecosystem this summer with Touch Surgery Aide, an AI-native surgical computing platform designed to bring real-time artificial intelligence capabilities directly into the operating room. Offering a preview at the Society of Robotic Surgery 2026 Annual Meeting in July, the Touch Surgery Aide combines computer vision, multimodal AI, and accelerated inference to process surgical video and procedural context in real time. This convergence of digital technologies enables multiple AI applications to run simultaneously during a live procedure to process and immediately act on surgical data. Touch Surgery Aide aims to bring intelligent, increasingly automated support in the OR and is designed to continuously learn and evolve, according to Medtronic.
“Real-time AI marks a new frontier for healthcare,” Medtronic Chief Technology Officer Jim Peichel stated during Touch Surgery Aide’s unveiling, “it is a fundamental evolution in how surgery is supported, moving from technology that assists the surgeon’s hand to technology that helps teams work smarter, move faster, and deliver greater impact in the moment.”
Johnson & Johnson MedTech is working toward the same objective by leveraging NVIDIA Corporation’s open-source, GPU-accelerated software capabilities. The tech giant’s Medical Physics Simulation framework—embedded within its Isaac for Healthcare platform—helps medical robotics developers model anatomy-device interaction, generate hard-to-capture scenarios, test in silico, and train or evaluate medical robotics policies before hardware-heavy testing.
Specifically, Johnson & Johnson MedTech is using NVIDIA’s Medical Physics Simulation and a Cosmos-based foundation model to construct digital twins of its endoluminal MONARCH platform for urology, simulating complex anatomy and kidney stone scenarios.
CMR Surgical, Moon Surgical, and Rob Surgical are adopting NVIDIA’s healthcare-specific AI tools as well to accelerate workflows. Perioperative intelligence/orchestration developer PeritasAI, meanwhile, is building a physical AI platform for surgical operations that integrates robots and multi-agent intelligence to sense, coordinate, and act in real time. In addition, Proximie—another physical AI platform architect—is creating vision-language models that combine operating room images with intraoperative video. These models power real-time AI agents that provide surgical insights and orchestration across the surgical pathway.
“Some of the richest and most complex data in healthcare is generated in the OR, yet much of this intelligence has historically remained underused and disconnected,” Proximie Founder/CEO Nadine Hachach-Haram explained. “At Proximie, we have spent years building the infrastructure to capture and structure this precious intraoperative data. By combining Proximie’s ambient surgical intelligence with NVIDIA AI technologies, we are building ORs capable of anticipating needs, supporting clinical teams, and continuously learning from every procedure.”
GE HealthCare is supporting clinical teams, too, through its collaboration with Mass General Brigham. Announced in September, the pair will investigate the use of generative AI in radiation oncology to help clinicians retrieve and interpret patient information during treatment planning and delivery.
The organizations particularly will explore building a generative AI tool that combines structured clinical data and unstructured information, including clinical notes, medical images, and treatment plans. The tool would enable radiation oncology teams to quickly retrieve relevant patient information and generate concise summaries of patient treatment journeys.
Both GE HealthCare and Mass General Brigham aim to incorporate the proposed AI technology into the former’s Intelligent Radiation Therapy (iRT) software, which is designed to coordinate radiation therapy workflows and reduce manual handoffs. The research could also inform future workflow capabilities involving MR, CT, and theranostics.
The collaboration and resulting generative AI tool will likely fall under the auspices of GE HealthCare’s new Global Chief AI Officer Rodolphe Katra. Appointed in September, the former Medtronic executive is tasked with advancing the company’s lead in FDA-authorized, AI-enabled medical devices, a portfolio that currently numbers 134.
“Rodolphe will help us take our AI strategy to the next level—connecting devices, software and data to deliver better insights, more longitudinal understanding and more precise care,” GE HealthCare President/CEO Peter J. Arduini stated. “At GE HealthCare, our goal is simple: build AI-enabled solutions that are making a difference for patients, clinicians and health systems.”
That goal resonates with Philips N.V. as well. Its Spectral CT Verida system, which received FDA 510(k) clearance in April, is billed as the world’s first AI-powered detector-based spectral computed tomography system that integrates across the AI imaging chain, from acquisition to reconstruction.
Verida incorporates Spectral Precise Image technology with a NanoPanel Precise dual-layer detector that provides intrinsic noise reduction, combined with an AI-based reconstruction engine that allows clinicians to customize image de-noising according to their preferences.
With the ability to reconstruct 145 images per second, Verida can render a complete exam in under 30 seconds and handle up to 270 exams a day. By capturing high- and low-energy data in a single acquisition, Verida delivers both conventional and spectral imaging results simultaneously, eliminating the need for additional scans or existing workflow changes. This capability enables tissue characterization and material differentiation without requiring advance image selection parameters or repeat image acquisitions, Philips claims.
Verida is designed for diagnostic imaging across radiology, interventional radiology, cardiology, and oncology treatment preparation, including radiation therapy planning.
“By combining always-on spectral imaging with AI-powered reconstruction, Verida enables clinicians to see more, first time right, supporting faster, more informed decisions and expanding the role of CT across clinical pathways,” Dan Xu, CT business leader at Philips, said upon Verida’s FDA clearance.
By enabling faster, more informed clinical decisions, AI-powered technologies such as Verida are helping pave the way for safer, more timely and personalized patient care. Philips’ 2026 Future Health Index U.S. Report underscores these benefits: 58% of clinicians reported improved workflow efficiency, while 54% saw faster decision-making. AI is augmenting clinical judgement rather than replacing it, enabling clinicians to make decisions more quickly with greater confidence and stronger collaboration across care teams.
“AI is no longer a future promise in healthcare. It’s beginning to change how care is delivered day to day. Not long ago, much of the focus was on isolated clinical applications—for example, how AI could diagnose cancer earlier or improve a single moment in care. Today, we’re seeing a broader shift: AI is becoming part of everyday care delivery, helping clinicians make better use of time, information, and expertise to better support patients,” the Future Health Index Report states. “Integrated AI could help care teams move toward more coordinated, proactive and personalized care—supporting earlier intervention, better patient experiences and improved outcomes. Today, AI has a clear role in healthcare. The next step is integrating it responsibly into care, so that together we can deliver better care for more people.”
A goal that would make ELIZA proud.
