3D anatomical digital twins: New generation virtual models to navigate robotic partial nephrectomy.

Objective 3D virtual models have gained interest in urology, particularly in the context of robotic partial nephrectomy. From these, newly developed "anatomical digital twin models" reproduce both the morphological and anatomical characteristics of the organs, including the texture of the tissues they comprise. The aim of the study was to develop and test the new digital twins in the setting of intraoperative guidance during robotic-assisted partial nephrectomy (RAPN). Patient and Methods The production path of the 3D model-digital twin of an organ begins with a phantom of virtual elements, including the kidney's parenchyma, vessels, tumour and collecting system. Textures are created from intraoperative robotic surgery images using machine learning algorithms. The result is a 3D model - digital twin that replicates the organ's shape and appearance. Two surgeons, one experienced and one young, used both the standard 3D model and the digital twin in four surgical phases: identifying the organ and its boundaries, dissecting the vascular pedicle, isolating the neoplastic lesion and identifying the renal pelvis and ureter. Results 4 patients, 2 per each surgeon harbouring a low and intermediate complexity (PADUA 6 and 8) renal masses respectively, underwent RAPN. From the assessment made by the surgeons at the end of each procedure, the 3D digital twin models were found to be superior to their standard counterparts both in terms of concordance with real anatomy and in usefulness to guide the identification of the tumour, vascular pedicle and ureter, while they did not demonstrate significant advantages in identifying the kidney and its margins. Conclusions The new 3D digital twin models represent a step forward towards the personalization of virtual reconstructions. Approaching real anatomy more closely, they offer the surgeons a perceived higher degree of concordance with the intraoperative environment, making it easier to identify the structures of interest during the surgical procedure.