Medtronic Unveils Touch Surgery Aide to Power Real-Time AI in the OR

The platform functions as the central engine for Medtronic's existing Touch Surgery ecosystem, which currently serves over 1,500 hospitals. By integrating NVIDIA’s Holoscan, CUDA, and TensorRT technologies, Touch Surgery Aide utilizes computer vision and multimodal AI to assist surgical teams in real time. The system processes video and contextual data, allowing multiple AI applications to run simultaneously during live operations.

The first application cleared for use on this platform is the Instrument Exit Point (IEP), which recently secured FDA approval for the Hugo robotic-assisted surgery system. IEP utilizes computer vision to notify surgeons when instruments move outside the visible field of view. This feature marks a strategic shift for the company, moving beyond passive tools to active, AI-driven assistance that aims to improve workflow predictability.

Medtronic plans to expand the platform’s utility to include laparoscopic surgery while continuing to scale the Hugo system's global footprint. With the Hugo platform already utilized in tens of thousands of procedures across more than 35 countries, the company views this compute-heavy approach as a generational shift toward embedded connectivity in surgical environments.