Siemens Healthineers Collaborates with NVIDIA on Healthcare AI

October 21, 2026NVIDIA and Siemens Healthineers have announced a strategic collaboration to explore AI-enabled technologies that could enhance the patient experience, support personalized treatment, streamline radiotherapy workflows, and expand access to care.

Enabled by unprecedented advances in radiotherapy and computational power, this next chapter in the organizations’ ongoing relationship is exploring how artificial intelligence could support treatment planning and delivery. This includes automating clinical tasks with AI agents, simulating care virtually, and harnessing instantaneous, real-time data processing during live radiotherapy sessions.

If successful, products and services created through the collaboration could one day address some of the most persistent challenges facing the industry while allowing for safer, less stressful, and more accessible treatment for patients across the globe.

Building new tools to solve radiation’s biggest barriers

As cancer diagnoses rise, the need for innovative solutions to systemic constraints has reached an all-time high. Over 21 million people were diagnosed with cancer globally in 2024, and that number is expected to rise to 35 million by 2050. As demand grows, healthcare systems also face increasing pressure to deliver care while managing chronic workforce shortages, mounting financial constraints, and increased operational complexity.

At the same time, resource-intensive treatments like adaptive radiotherapy are generating vast amounts of data and adding complexity to clinical workflows. While these solutions represent some of the most advanced forms of radiation therapy available today, they remain best suited for complex cancer cases within healthcare systems that can handle the increased demands on their finite resources.

To make personalized, adaptive radiation therapy the new global standard in a manner that is both logistically and economically feasible, healthcare systems must overcome several critical barriers:

• Care teams face increasing workloads and workforce pressures.
• The ability to model treatment performance as anatomy changes is limited.
• Since every patient is different, personalized treatment planning often requires additional time and expertise, which limits treatment volume.
• Making adaptive radiotherapy universal demands more clinicians than will probably ever exist.

Addressing these challenges requires new ways of combining clinical expertise, data, simulation, and AI into more efficient, automated, and integrated solutions without adding to the existing burden on providers. This is where the healthcare AI collaboration between NVIDIA and Siemens Healthineers has the potential to streamline clinical workflows, alleviate pressure on care teams, and transform the patient experience.

How healthcare AI can make adaptive radiotherapy more accessible

NVIDIA and Siemens Healthineers have identified three primary areas in the adaptive radiotherapy space for initial development.

Agentic AI in treatment planning

In adaptive radiotherapy, clinicians must weigh numerous factors, including target coverage, organ sparing, treatment efficiency, and changing patient anatomy, while the tools that enable them to do so require more hands-on time from a larger number of specialists.

Agentic treatment planning offers an exciting opportunity to simplify the process. Through a collaborative co-planning approach, clinicians could work alongside AI agents in the future to evaluate treatment options, perform calculations, and highlight insights. Additionally, AI healthcare agents could speed up care delivery by automating tasks, coordinating workflows, or interacting with clinicians conversationally throughout a patient’s visit.

For example, rather than manually building and comparing treatment plans, a radiation oncologist could describe their clinical objectives and collaborate with an AI agent to explore potential strategies. Powered by NVIDIA Nemotron reasoning models and NVIDIA CUDA-accelerated dose simulations, these systems could evaluate candidate treatment plans in seconds and highlight tradeoffs for clinician review, with final plan review and approval remaining firmly in expert human hands.

In simple terms, AI could soon revolutionize the treatment room in the same way it has revolutionized the way we plan, learn, and create in our personal lives. With the support of healthcare AI agents tailor-made for adaptive therapy, the goal is to provide clinicians with more capacity to make the best decisions and provide more personalized, patient-centered care.

Advanced patient simulation

Artificial intelligence technologies like AI contouring outline targets and critical structures automatically to save clinicians up to an hour of work per treatment plan. But what if that power to streamline planning was harnessed from start to finish?

Advanced patient simulation promises to revolutionize treatment planning through patient-specific digital twins. Imagine if providers had a virtual patient they could use to create, simulate, and revise treatment plans without the patient present. With the support of the NVIDIA Isaac for Healthcare development platform, future technologies could simulate how radiation dose is delivered to tumors and surrounding healthy tissue before treatment begins, allowing care teams to evaluate different approaches before a dose is delivered.

Unlike static representations of the patient used during initial planning, a patient-specific digital twin could evolve alongside the patient. New imaging acquired throughout treatment would continuously update the model and reflect anatomical changes as they occur. Care teams could use that information to evaluate whether treatment remains aligned with the original plan and determine when adjustments may be necessary.

These patient-specific simulation environments using NVIDIA Omniverse could allow care teams to evaluate plan quality, understand dose behavior, explore treatment scenarios, and optimize treatment strategies to save time, provide greater confidence, and support more personalized cancer treatment.

Real-time adaptive computing in the treatment room

Treatment is dynamic. Patients move, anatomy changes, and tumors can shift position. Yet radiation therapy requires extraordinary precision to deliver the prescribed dose while protecting healthy tissue. Maintaining that precision can require significant coordination, clinical oversight, and computational resources, particularly in adaptive treatment environments.

Using NVIDIA Holoscan, which brings together sensor processing, networking, and ultra-low-latency computing with NVIDIA IGX, future solutions from Siemens Healthineers could process imaging, sensor, and treatment data as they’re generated. Instead of depending on separate observations and decisions, future radiotherapy delivery systems might develop a continuous, real-time understanding of the treatment for providers to leverage.

This level of awareness could fundamentally change how adaptive care is delivered. Subtle patient movement that may once have required manual assessment could be identified and evaluated in real time. Clinicians would gain a clearer understanding of how those changes affect treatment, opening the door to more responsive approaches without introducing additional workflow complexity.

Our commitment to enhancing the patient experience

As this collaboration explores the future of healthcare AI in adaptive radiation, Siemens Healthineers continues to unveil major developments in cancer care today.

A recent example is the Accela* radiotherapy system, powered by NeoArc.* Introduced at ASTRO in September 2026, Accela is engineered around NeoArc (Dynamic Modulated Arc Therapy - DMAT), a novel treatment technique designed to help clinicians overcome traditional tradeoffs between treatment complexity, delivery speed, and workflow efficiency. The system supports imaging-to-treatment workflows in as little as 90 seconds, enables multi-site SBRT treatment delivery in as little as 60 seconds, and can acquire CBCT images in a single 15-second breath hold.

Advances in treatment delivery platforms like Accela and emerging AI technologies represent an opportunity to evolve adaptive radiotherapy further. If we can pair faster, more capable treatment systems with agentic AI, advanced simulation, and real-time computing, we hope to build a world without fear of cancer, where personalized cancer care is not only possible, but practical and scalable for providers and patients everywhere.

*510(k) pending. Not available for sale in any market. No guarantee of commercialization or feature availability.

• Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, Jemal A. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024 May-Jun;74(3):229-263. doi: 10.3322/caac.21834. Epub 2024 Apr 4. PMID: 38572751.