Despite the medical and surgical progress of the last two decades, the selection of candidates for liver surgery remains based on old principles and insufficiently sensitive to fine-tune the gesture to patient-specific characteristics and make almost zero risks of postoperative liver failure (PLF) and death. It is therefore necessary to develop new tools that will make possible to predict the evolution of the postoperative portocaval gradient (difference of pressure between portal vein and vena cava), a well-known major risk factor for PLF. Hemodynamic modeling of the human liver during surgery will represent the purpose of this work in order to help the clinicians in their patient's selection and anticipation of postoperative risk. The aim is to develop and validate an hemodynamics mathematical model to predict the evolution of the portocaval gradient in three surgical situations of increasing complexity: portal modulation by embolization, hepatectomy, and small partial graft liver transplantation. The endpoints will be the estimation of the intraoperative post-procedural portocaval gradient and comparison of the estimated portocaval gradient with that measured at the end of the procedure. This pressure differential is performed before parietal closure, after surgery.