659 Enhanced surgical guidance via real-time 3D fusion

Dr. Marco Mezzina

Minimally invasive and robotic-assisted surgeries demand accurate visualization and reliable tracking of anatomical structures, yet surgeons often lack seamless real-time access to preoperative imaging such as CT scans. This limitation can compromise decision-making and increase intraoperative uncertainty, especially in complex procedures. To address this problem, we developed a real-time fusion system that integrates patient-specific 3D anatomical models with live intraoperative video. The solution employs a complex processing pipeline combining stereo depth estimation, optical flow, and segmentation algorithms, all optimized for low-latency performance on dedicated hardware. By tracking anatomy in real time and overlaying preoperative meshes onto the endoscopic view, the system provides surgeons with intuitive, context-rich guidance. The key innovation lies in the integration of multiple AI-driven algorithms running in parallel within a clinically acceptable latency window (<75 ms), enabling reliable real-time anatomical overlays. This enhances surgical navigation and has shown promise in improving outcomes during robotic-assisted partial nephrectomie
Technological Innovation, Artificial Intelligence, Robot-assisted partial nephrectomy (RAPN)

See more at: https://vattikutifoundation.com/videos/