Proximie and NVIDIA Collaborate to Power AI-Driven Surgical Intelligence and Next-Generation Robotic Operating Rooms
Proximie, a global leader in intelligent operating room (OR) software, is significantly expanding its capabilities by leveraging the advanced AI infrastructure and foundation models developed by NVIDIA. This collaboration marks a pivotal step toward the realization of fully intelligent, data-driven surgical environments, where artificial intelligence and robotics seamlessly integrate into clinical workflows. At the center of this partnership is Project Rheo, a newly launched initiative by NVIDIA designed to accelerate the development of healthcare robotics and intelligent clinical ecosystems. By combining Proximie’s extensive real-world surgical data with NVIDIA’s powerful AI platforms, the partnership is laying the groundwork for scalable physical AI systems that can transform operating rooms across the globe.
Proximie has long positioned itself as the operating system for the modern OR, and its Smart OR platform continues to evolve as a central hub for surgical intelligence. Built on years of data capture and workflow optimization, the platform is now being enhanced through NVIDIA’s accelerated computing stack, enabling faster data processing, deeper insights, and more advanced AI model training. In return, Proximie contributes one of the most comprehensive repositories of surgical data available today, offering invaluable real-world insights that inform the next generation of NVIDIA’s AI models.
At the core of Proximie’s technology is its Intelligence Suite, an ambient AI infrastructure layer designed specifically for surgical environments. This system continuously captures and processes vast amounts of intraoperative data from live operating rooms across hundreds of healthcare institutions worldwide. The data collected includes high-resolution intraoperative video, workflow activity snapshots, instrument usage patterns, and detailed procedural context. Unlike traditional healthcare data systems, which often rely on fragmented or retrospective data, Proximie’s platform operates in real time, creating a continuous stream of structured and actionable intelligence.
This rich dataset feeds directly into NVIDIA’s AI ecosystem, enabling the development of increasingly sophisticated models that can interpret and act upon surgical workflows. With the support of the open NVIDIA Cosmos platform, Proximie’s data is transformed into AI-ready datasets at scale. Cosmos allows for advanced synthetic data generation, a critical capability in healthcare where real-world data can be limited by privacy, variability, and regulatory constraints. By augmenting real surgical data with high-quality synthetic data, the collaboration ensures that AI models are robust, diverse, and capable of generalizing across different clinical scenarios.
A key component of this effort is Cosmos-H, NVIDIA’s specialized foundation model designed to simulate surgical environments and generate synthetic datasets tailored to operating room use cases. Cosmos-H enables the creation of realistic surgical scenarios that can be used to train AI systems without the limitations of traditional data collection. These datasets are instrumental in developing advanced vision-language models capable of understanding complex surgical procedures in real time.
These vision-language models represent a major leap forward in OR intelligence. By analyzing both visual inputs and contextual data, they can monitor surgical activity as it unfolds, recognize critical milestones during procedures, and provide real-time insights to clinical teams. More importantly, these models can trigger physical actions within the OR environment, bridging the gap between digital intelligence and physical execution.
This is where the concept of physical AI becomes transformative. Through Project Rheo, AI-driven insights are not limited to observation and analysis—they extend into action. The collaboration envisions a new generation of robotic systems, including humanoid assistants, that can actively support surgical teams. These robots could prepare surgical instruments, retrieve necessary equipment, and assist in maintaining workflow efficiency, all based on real-time AI-driven understanding of the procedure. In this integrated system, one layer of AI continuously observes and interprets the surgical environment, while another layer executes tasks, creating a synchronized and highly responsive operating room.
For healthcare providers, the implications of this technology are both immediate and long-term. Hospitals using Proximie’s platform have already reported significant improvements in operational efficiency. OR productivity has increased by as much as 24%, enabling facilities to perform up to 300 additional procedures per operating room each year. These gains are largely driven by improved visibility into surgical workflows, allowing for more accurate predictions of procedure durations, better scheduling, and optimized resource utilization.
Project Rheo builds on these gains by introducing adaptive intelligence into the OR. Rather than simply optimizing existing workflows, the system continuously learns from every procedure performed. Each surgery contributes new data, which is then used to refine AI models and improve future outcomes. Over time, this creates a self-improving system where the standard of care is constantly elevated through data-driven insights.
Nadine Hachach-Haram, Founder and CEO of Proximie, emphasized the significance of this evolution in surgical intelligence. She noted that operating rooms generate some of the most complex and information-rich data in healthcare, yet much of this data has historically remained underutilized. Proximie’s mission has been to unlock this potential by building the infrastructure needed to capture, structure, and analyze intraoperative data at scale.
Through Project Rheo, this vision is being realized in new and impactful ways. Real-world surgical data is now being used to train AI systems that can truly understand the dynamics of the operating room. These systems are not only capable of providing insights but are also enabling the development of next-generation healthcare robotics, both digital and physical. By combining Proximie’s ambient intelligence with NVIDIA’s AI technologies, the collaboration is paving the way for operating rooms that can anticipate clinical needs, support surgical teams in real time, and continuously improve through learning.
One of Proximie’s key differentiators is its ability to capture both intraoperative and wide-field OR video simultaneously. This capability extends beyond traditional operating rooms to include endoscopy suites and catheterization labs, creating a comprehensive and diverse dataset that spans the full spectrum of procedural environments. This breadth of data provides AI systems with the contextual understanding needed to accurately interpret surgical workflows across different specialties and settings.
As a result, Proximie has built one of the largest and most clinically diverse repositories of surgical intelligence in the world. This dataset serves as a foundational asset for training AI models that are not only accurate but also adaptable to real-world clinical variability. By continuously feeding this data into NVIDIA’s AI ecosystem, the partnership ensures that both companies remain at the forefront of innovation in healthcare technology.
Looking ahead, the integration of AI and robotics into operating rooms represents a fundamental shift in how surgical care is delivered. Intelligent ORs will no longer be passive environments but active participants in the surgical process. They will provide real-time guidance, automate routine tasks, and enable clinicians to focus on the most critical aspects of patient care.
In this emerging paradigm, the collaboration between Proximie and NVIDIA stands as a powerful example of how technology can be harnessed to improve healthcare outcomes. By combining advanced AI infrastructure with real-world clinical data, they are not only enhancing current surgical practices but also shaping the future of medicine. The intelligent operating room is no longer a distant vision—it is rapidly becoming a reality, driven by innovation, data, and the seamless integration of digital and physical intelligence.
Source link: https://www.businesswire.com
