To describe a new Augmented Reality PC based software of 3D segmentation from CT scan. To present a new 2D / 3D fusion software of fluoroscopic images and segmented volumes. To teach and to train residents to facilitate percutaneous renal puncture.

The main problem in PCNL is the precise localization of organs and lesions in order to identify the best percutaneous trajectory. We developed an Augmented Reality (AR) system and real time registration software that allows to compute 3D segmentations of body skin, iliac crest, 11th-12th ribs, spine, kidney, renal cavities and stones from CT scan examination, and then choose the optimal PCNL trajectory. During the procedure, the projection of the segmented 3D volumes and computed trajectory (skin target, length and angle of puncture) on the patient body shows in real time the target and the best PCNL way. Furthermore, the fluoroscopic 2D data are fused on-line with the 3D segmented structures, allowing the surgeon to determine the precise position of the surgical needle and correct its trajectory. This procedure leads to a drastic decrease of radiations and also allows teaching and training residents on kidney phantoms in order to facilitate the percutaneous access in operating room.