Smart surgery through AI-based assistance systems

As part of the research project, a digital OR assistant is to be developed, which will make it possible to relieve the staff in the OR by monitoring instruments and also to acquire data for planning more efficient and hygienically optimized ORs. Multiscale AI-based image processing for object recognition and coordinate regression will be central to the development of the OR assistant, which will make it possible to track relevant objects in the room with cost-effective and easy-to-install cameras and deliver robust results even in the event of occlusions. Central research questions include in particular network conditioning with synthetic data based on detailed virtual 3D environments and the development of network architectures in combination with state estimation methods. The recognition algorithm is used to monitor all tools in use and documents each individual tool in terms of duration of use, type of use, user and movement profile. All this information enables the creation of reliable statistics and the automated preparation of a surgical report, intelligent surgical planning or the design of gripper systems for automated instrument application. The project is funded by the Young Investigator Grands funding line of Leibniz Universität Hannover.