XiangshuTech: Spatial AI for the Operating Room

Field notes

XiangshuTech frames surgery as a perception-and-coordination problem: preoperative imaging, intraoperative deformation, tissue semantics, and ultrasound are intended to meet inside the surgeon’s view. The homepage names a LIBR/LIBR+ deformation-correction layer, mixed-reality navigation, markerless tracking, and collaborative robotics rather than presenting AI as a generic assistant. Its cited performance figures distinguish liver, breast, and phantom testing, with the breast result listed at 4.2 mm and the phantom result at 1.7 ± 0.8 mm.